The comprehensive database of AI deployments. Every real-world AI deployment in any country, tracked & sourced 24/7.
We track AI deployments in every country on earth, grouped here into six regions. Tap a region to filter the record, or name a specific country in the filters below.
Read this often? Tell Google to show you more of it — set the Institute as a preferred source and our entries surface higher in your results.
Add us as a preferred source on GoogleEvery entry is a named organisation putting AI to work in a real job — with the evidence and the rules forming around it — each one sourced and dated. Open an industry to read its entries, or use the filters below.
Every entry is filed four ways — what it is, the industry, where in the world, and the job it changes.
Ministry of Justice (Egypt) — Ministry of Justice said it will start AI court transcription on 1 October, using speech-to-text for Egyptian dialects. Spoken proceedings become automated transcripts of hearings under judicial digitisation.
Delhi Pollution Control Committee — Delhi Pollution Control Committee's Integrated Command and Control Centre at Shastri Park, opened by Chief Minister Rekha Gupta, uses AI pollution monitoring, CCTV and GPS on 700-plus construction sites over 500 square metres. Body cameras feed the AI-led hub; 78 sub-division teams will enforce from there this winter.
Cambridgeshire Constabulary — Cambridgeshire Constabulary is trialling Citizen Assist, an AI crime reporting tool from the National Police Chiefs' Council's Digital Public Contact programme. Residents describe non-emergency offences conversationally and check the finished account before sending. Assaults, threats, damage and theft come first; wider adoption of AI-assisted reports depends on results.
Ministerio de la Presidencia, Justicia y Relaciones con las Cortes; Tribunal Supremo — Spain's Ministry of the Presidency, Justice and Relations with the Courts deployed AI in the Supreme Court, letting First (Civil) Chamber legal teams extract parties, issues, cited provisions and arguments from appeals in seconds. The chamber logged 7,279 new cases in the first half of 2026; 100-plus staff trained in July. This AI case analysis runs on isolated, state-owned open-source systems.
Moka City (Tochigi Prefecture) — Moka City in Tochigi Prefecture put a generative AI chatbot for residents online, supplied by SDT. Built from roughly 4,700 city pages, AI answers on municipal services — paperwork, taxes, childcare, health, welfare, rubbish — run 24 hours, 365 days a year.
Glavgosexpertiza of Russia — Glavgosexpertiza of Russia started trialling AI design review, a large-language-model assistant for its experts. Before formal examination, project files get scanned for likely defects and gaps, delivering AI checks for construction as a preliminary report.
Tribunal de Justiça da Bahia (TJBA) / Tribunal de Justiça de Mato Grosso (TJMT) — Brazil's Bahia state court (TJBA) began installing Sistema Hannah, the Mato Grosso state court's (TJMT) court AI tool, on 21 September and trained 2nd Vice-Presidency magistrates and staff on 23 September. Hannah scans case files, tests appeals bound for the STJ or STF against a 14-criterion admissibility map and drafts a structured report for the ruling. In Mato Grosso this admissibility assistant closed over 15,000 cases, shrinking typical turnaround from 23 days to two.
Municipalidad de Punta Arenas — Punta Arenas is the first Magallanes municipality to adopt Revi, the Chilean Chamber of Construction's AI for building permits, joining 17 councils nationwide. Norman Revi helps Dirección de Obras Municipales reviewers draft observations and verify paperwork, while Clara Revi screens applicants' files beforehand. Mayor Claudio Radonich expects roughly 30% quicker permit approval; CNEP puts the national average at 120 to 126 days.
Comptroller and Auditor General of India — The Comptroller and Auditor General of India and the Digital India BHASHINI Division released CAG-BHASHINI Anuvaad for CAG's AI translation of audit reports. Built for consistent terminology, data sovereignty and restricted access, the translation tool keeps human oversight.
Dewan Perwakilan Rakyat (DPR RI) — Indonesia's House of Representatives (DPR RI) unveiled PILAR Ber-AKSI, pairing the PILAR-AI Parliament framework with the AKSI platform for AI-assisted legislative drafting. AKSI links regulations and provisions; drafters and members keep judgement.
Kementerian Dalam Negeri (Ministry of Home Affairs, Indonesia) — Ministry of Home Affairs is widening AI welfare verification from 43 to 258 regencies and cities, minister Tito Karnavian said. Population records are matched with asset data from eight ministries and agencies.
Provincia del Neuquén (Dirección de Fauna, Ministerio de Seguridad) — Neuquén province tracks native fauna through 170 cameras feeding an AI wildlife monitoring hub, Argentina's first such provincial outlay in 20 years. Artificial intelligence keeps the hub running 24 hours a day under three supervisors, with automatic movement alerts and stored video. Another 50 units are due before December, building a 300-camera wildlife surveillance network.
Municipalidad de San Miguel de Tucumán — San Miguel de Tucumán's municipality has launched Migue, an AI municipal assistant answering residents' paperwork questions 24 hours a day in Argentina. Linked to the CiDiTuc digital-citizen platform, the website and Telegram, Migue handles plain-language queries about driving-licence bookings, business permits and traffic penalties, redirecting people to official channels when stuck. Mayor Rossana Chahla unveiled the city chatbot at City Day, backed by Bloomberg Philanthropies' City Data Alliance with Johns Hopkins University coaching.
Sistema Nacional de Atención a Emergencias y Seguridad 9-1-1 — The Dominican Republic's 9-1-1 emergency and security system showcased AI emergency response tools, first-response drones and a Boston Dynamics robot at President Luis Abinader's LA Agenda Semanal. A 9-1-1 call can trigger geolocation, dispatch and a VAPRE-M2 drone streaming live images, with city cameras and NVIDIA-hosted AI offered as decision support. The robot is to assist agencies in high-risk emergency dispatch work.
US federal government (White House) — America.gov, the Trump administration's AI government chatbot built on Google's Gemini and xAI's Grok, handles US questions about passports, campsites and Social Security cards. Replies draw on roughly 29,000 official sites and currently redirect users; 2027 upgrades promise AI help with public services like finishing applications online.
Buk-gu Office, Gwangju — Buk-gu in Gwangju answers after-hours complaints with an AI call-bot and gives staff a generative platform for memos, plans and press notes. The automated complaints line routes callers to information, the right department or emergency transfer, sending blazes, crashes or unrecognised requests to the duty officer. A year-end review follows.
Kementerian Hukum (Ministry of Law) — AI regulatory synchronisation now lets Indonesia's Ministry of Law (Menteri Hukum) scan its entire legal corpus for clashes among hundreds of thousands of statutes and rules in force nationwide. Minister Supratman Andi Agtas said the tool spots duplicate clauses, contradictions and overlapping authorities that no human team could review at that scale. Institut Teknologi Bandung (ITB) is helping the ministry build the underlying AI-assisted legal harmonisation platform.
Thanh tra TP HCM (Ho Chi Minh City Inspectorate) — Staff at HCMC Inspectorate built their own AI complaint handling tool, now sorting complaints, denunciations and petitions in-house. Offline operation keeps informants unidentified, pulling from 456 digitised articles across 12 legal documents and 52 templates — AI-assisted petition processing already running agency-wide.
Ministry of the Interior and Safety — The Ministry of the Interior and Safety plans roughly 940 billion won for 50 AI disaster response projects by 2030, forecasting hazards and recommending evacuations and controls. Busan, Yeongcheon and Inje County pilot heavy-rain and typhoon support for field officials after an 8 September kickoff. Effectiveness checks follow early next year for this AI-driven emergency management.
Taipei City Revenue Service — Taipei's revenue service offers an AI tax assistant on Google NotebookLM, answering round-the-clock queries on ombudsman requests, objections and local levy exemptions with citations. People are warned not to type identities or case particulars, and individual disputes stay with staff ombudsmen rather than the tax chatbot.
Ministerio de Medio Ambiente y Recursos Naturales (MARN) — El Salvador's Ministry of Environment and Natural Resources has formalised SIMACC, a climate monitoring platform in force since 7 September that pools geo-environmental records from public bodies and local governments. The ministerial agreement permits automated analysis, predictive modelling and AI on environmental data, with outputs serving only as technical input for officials' decisions.
GovTech (Government Technology Agency of Singapore) — GovTech-built tools now run AI penetration testing on roughly 2,000 public-sector systems, and a second model reviews application code for flaws. Hunting intruders proactively became the priority after the UNC3886 espionage campaign; no go-live date for this automated security testing was given.
Ministry of Digital Development and Information / GovTech — SG Translate 2.0 from the Ministry of Digital Development and Information and GovTech blends large language models with neural machine translation, giving departments AI translation in four official tongues. Portals and e-services can plug in this multilingual service engine; Polyglot, which localises state web pages, already uses it.
Austrian Federal Chancellery (Bundeskanzleramt) — Austria's Federal Chancellery will embed AI for civil servants inside ELAK, the Bund's digital case-file system. Phased rollout of KI im ELAK begins within weeks and ends by late 2026, letting officials query records, condense paperwork and get answers without switching tools. Sovereign LLM-as-a-Service capacity at the Federal Computing Centre (BRZ) hosts this public-sector AI, and GovGPT already reaches around 20,000 federal employees.
Ministério da Fazenda — Brazil's Ministry of Finance already runs AI to detect illegal betting sites, linking gambling pages with the payment channels behind those pages. Finance Minister Dario Durigan unveiled the plan on 25 September with the Federal Revenue Service, the Federal Police and watchdog Anatel, after a provisional measure outlawed betting operators. Platforms still live from 6 October face blocking in this gambling enforcement drive.
State Tax Service of Kyrgyzstan — Kyrgyzstan's State Tax Service has added Aisalyk, an AI tax assistant for taxpayers' queries, to its relaunched website. The rebuilt portal aims to ease dealings with the authority for households and firms, and shows budget settlement balances on the front page without sign-in, alongside the virtual tax adviser.
Ondo State Government — Ondo State Government runs AI in civil service payroll, staff files and retirement through Ondo AI, unveiled by Governor Lucky Aiyedatiwa at the Governor's Office Cocoa Conference Hall, Akure. Forerunner Ondo Pay, an AI-enabled HR platform, has paid 2,802 recipients N9.7 billion in pensions and gratuities via automated pension processing.
Gwangjin-gu Office, Seoul — Gwangjin-gu in Seoul now runs AI CCTV search to trace missing children and older people with dementia. Given a photo or description, the software checks an hour of recordings from 100 cameras in roughly ten minutes, mapping movements over time for police. Officers received manuals and coaching for the automated video search.
Social Insurance Bank (SVB) of Curaçao — Curaçao's Social Insurance Bank (SVB) applies AI healthcare fraud detection to flag incorrect, unnecessary or illegitimate billing. Analysts and inspectors verify flagged cases via site checks and patient interviews, pairing automated anomaly detection in medical billing with human review rather than automation alone.
Tribunal Superior Eleitoral (TSE) — Brazil's Superior Electoral Court (TSE) has put ChatVote, an in-house AI voter assistant, on the TSE Portal homepage and the e-Título app. Voters can ask in plain language about the 2026 elections: polling day, hours and locations, excusing a missed vote, poll-worker duties and election resolutions. Bought-in chatbots served since 2020; this election chatbot is the first built by the court itself. Answer ratings, WhatsApp access and voice trials come next.
Prefeitura de Senhor do Bonfim — Senhor do Bonfim, Bahia, is readying BIA (Bonfim Inteligência Artificial), an AI tax assistant serving the Municipal Department of Finance and Budget. Residents will request IPTU, TFF and TLL bills, arrears lookups, payment slips, statements and certificates by message, voice note, call, video call or snapshot of paperwork. Web chat and council channels will host the municipal taxpayer chatbot.
Colorado Springs Police Department — Colorado Springs Police Department's Public Safety Communications Center has switched on Prepared by Axon to transcribe, translate and summarise calls and route non-emergency calls. S.A.R.A.H., an AI agent for non-emergency calls, now answers those lines first; trained call takers keep every 911 call. Automated call answering was tested from April 2026.
Tennessee Department of Commerce & Insurance — The Tennessee Department of Commerce & Insurance has rolled out Iris, an AI chatbot for professional licensing that answers round the clock for over 357,000 licensees in 26 regulatory programmes. Via the CORE system, in English or Spanish, the licensing assistant also looks up licensed individuals and businesses. A soft launch ran in August.
Bundesagentur für Arbeit (Federal Employment Agency) — Klara, an AI voice assistant government hotline tool for Germany's Federal Employment Agency (BA), now answers roughly 18 million SGB III calls yearly. She greets callers, logs identity and basic needs, and in four pilot regions settles routine unemployment-benefit questions unaided. This telephone-support automation aims to shrink hold times, forwarding harder cases to staff with details already logged.
Department of Information and Communications Technology (DICT) — The Department of Information and Communications Technology switched on eGov AI within the eGovPH app, a helper named Kuya A that brings AI in government services to confirmed account holders at no charge. The government AI helper reads scans, renders Philippine tongues, decodes paperwork and answers travel queries.
Nowon-gu Office, Seoul — Nowon-gu in Seoul trials OnAI, the Ministry of the Interior and Safety's generative AI for civil servants, from October. Officials will query rules, crunch data, draft papers and auto-generate minutes; 8 regional and 46 local authorities join. Nowon GPT already serves 1,150 of 1,584 staff in AI-supported administration.
Chuncheon City — Chuncheon City's second supplementary budget, passed by councillors, puts 1.647 billion won into ten projects including AI welfare checks. Power and telecoms readings drive AI-based check-ins, alerting case workers when someone living alone breaks routine, widening coverage from 200 to 500 people. CCTV detection of fly-tipping and about ten officials building bespoke tools complete the package.
Tax Committee of Uzbekistan — The Tax Committee of Uzbekistan is taking the AI Call Center nationwide: AI debt collection calls tell people owing tax their balance and deadline, no staff involved. In a Gallaorol trial on Jizzakh property-tax debtors, 52% paid in full at once after this robocall collection.
Comune di Argenta — The Comune di Argenta switched on Civia on 18 September, an AI assistant for council services fielding queries about forms, cut-off dates and what the town has on. Type the query as you would speak it — a nursery place, Saturday's listings, a copy of a certificate — and Civia replies, 24 hours a day, across languages. Justomezzo, whose trade is public-sector messaging, built the municipal AI chatbot with machine-learning work from Momentum Digital.
Nigerian Defence Academy — Chief of Defence Staff General Olufemi Oluyede has commissioned the Nigerian Defence Academy Command Centre in Kaduna, built around AI CCTV surveillance. The cameras should sharpen situational awareness and better shield cadets, staff and critical infrastructure. Oluyede urged the armed forces to adopt AI-assisted security monitoring faster as threats shift.
Immigratie- en Naturalisatiedienst (IND) — Two undisclosed chatbots, Robin and ChatIND, now handle AI asylum processing tasks for the Netherlands' Immigration and Naturalisation Service (IND), assisting roughly 230 workers with document summaries, translations and interview analysis since early 2026. Despite IND's stated transparency rules, neither tool appears in the country's official algorithm register. Critics among lawyers and researchers fear flawed outputs from AI-assisted asylum applications could distort decisions on refugee claims.
Governo de Goiás / Corpo de Bombeiros Militar de Goiás — Goiás has activated the Sistema Integrado de Monitoramento e Alertas de Goiás, an AI wildfire detection platform flagging probable forest-fire hotspots from satellite imagery. Officials claim two to four hours saved in operational response; several specialists previously sifted those pictures manually. AutomaGo IA co-built it. Satellite fire monitoring gains here carry no independent evaluation.
Regione Piemonte — Regione Piemonte has switched on Auxil.IA, a Csi Piemonte build putting AI for public administration into the hands of its workforce and town halls region-wide. Launched 17 September at the Grattacielo Piemonte alongside Paolo Zangrillo, Alberto Cirio and Csi Piemonte chief Pietro Pacini, it shortens files, weighs versions, writes up meetings, handles languages and parks everything in the Region's cloud. Councils may apply from 1 October, widening generative AI in local government.
Ayuntamiento de Ciudad Real — Spain's Ciudad Real City Council will buy a Municipal Artificial Intelligence Platform through a 447,700 euro, four-month contract. Staff are to apply AI for local government to resident enquiries, e-office casework and building permits, with the supplier configuring and linking these municipal AI tools.
Prefeitura de Salgueiro — Salgueiro, in Pernambuco, now answers residents on WhatsApp through SIA, a municipal AI assistant wired into the 1Doc service centre. Questions on public services, tax, urban licensing and the Municipal Ombudsman get replies at any hour, by text or voice note. The round-the-clock council chatbot passes specialist requests through 1Doc to the relevant department.
Government of Ecuador — Ecuador's government will install Edge, AI border surveillance from the United Arab Emirates, on the Colombia and Peru frontiers, President Daniel Noboa said. Edge should spot illicit activity, odd movements and cargo tampering; Noboa judged the technology better value than 50,000 soldiers. No price or schedule for the frontier monitoring AI has been given.
Trading Corporation of Bangladesh — Trading Corporation of Bangladesh (TCB) has built AI subsidy monitoring into digital operations that Commerce Minister Khandakar Abdul Muktadir opened on 17 September. Officials can follow sales and hand-outs area by area, a form of real-time distribution tracking, for roughly 7.9 million beneficiaries. Savings this fiscal year could reach Tk 3 billion to Tk 3.5 billion against last year.
Hellenic Court of Audit — Greece's Hellenic Court of Audit (Ελεγκτικό Συνέδριο) now runs an AI virtual assistant for judges, an AI-assisted judicial tool hunting through statutes and precedent to sort case files faster. Goal: shrink rulings from near 1,500 days to 600-650 days by end-2027. This came with a digital and energy revamp, moving filings and pre-contractual checks on public contracts fully online.
Tribunal de Contas do Estado da Paraíba (TCE-PB) — Tribunal de Contas do Estado da Paraíba has switched on Seixas, an AI for public-accounts auditing. Built jointly with Universidade Federal de Campina Grande, the system fields ordinary queries and hands back the exact wording underpinning every reply, sweeping 8.78 terabytes: 212,362 audit reports, 69,809 decisions, 16,443,401 other procedural acts. Automated drafting arrives at phase two.
Dubai Electronic Security Center — Dubai Electronic Security Center has shown SARAAB, an openly published AI deepfake detection model that watches a clip end to end, at GISEC Global 2026 in Dubai Exhibition Centre. The centre calls it Emirati-made and the first such tool from an Arab government body. Developer platforms will carry the artificial intelligence video forensics model for engineers and security staff.
Ministry of Defence of Kazakhstan — Kazakhstan's Ministry of Defence now runs AI in military conscription, with software proposing the best-fitting branch and type of service for each recruit from body measurements and schooling. These automated service-branch recommendations guide the commissions, which alone decide who is called up after reviewing medical examination results.
City of Munich and Munich Police — Munich police and city officials in Germany put over 50 AI cameras at around 35 Oktoberfest spots for AI crowd monitoring, catching dangerous crowding early. Fraunhofer Institute staff read density and movement from footage anonymously. A new hub lets police, fire, rescue and transport crews steer visitors through that crowd analytics.
Macao Meteorological and Geophysical Bureau (SMG) — Storm tracking in Macao just got smarter: the Meteorological and Geophysical Bureau (SMG) confirmed it now runs AI severe weather prediction tools across daily forecasting, alerting and hazard analysis. Ho Ion Sang, a lawmaker, prompted the disclosure through a written question to the bureau. Ties with Chinese meteorological agencies continue, and MAZU, the bureau's smart weather platform, is set for an upgrade toward AI-based weather forecasting alerts.
Committee for Emergency Situations and Civil Defense of Tajikistan — ARGUS, an AI early warning system from zehnlab.ai, Indian firm SkyServe and Tajikistan's Committee for Emergency Situations and Civil Defense, now guards the Pamir Mountains from orbit. A satellite GPU runs a computer vision model onboard, spotting floods, glacial lake bursts, landslides and snow slides, then sends short alerts instead of raw pictures. This orbital AI monitoring starts at Lake Pshart on 20 September 2026, later reaching Lake Sarez.
Prefeitura de Maringá — Prefeitura de Maringá has opened Mari, an AI assistant for council services reachable on WhatsApp. Residents describe a need in ordinary words and get guidance shaped to them, rather than learning how the city hall is organised. Instituto de Projetos Avançados para Cidades, Tecnologia e Administração and Secretaria Municipal de Logística e Compras run the chatbot within Maringá + Digital.
City of Cambridge — The City of Cambridge has switched on two AI tools for development approvals aimed at faster home construction. Archistar tests a proposed scheme against municipal requirements before anyone files, and the Cambridge Digital Twin renders the city in three dimensions so planners, developers and residents can see what is proposed. Since September 1 builders, architects and homeowners can put designs and requirements through this AI planning check.
Municipality of Asunción (911 System) — Paraguay's government is rolling out an AI-powered 911 emergency system in Asunción, backed by a USD 10 million project from South Korea's KOICA (Korea International Cooperation Agency). It covers intelligent video surveillance and a unified monitoring platform, running through 2030 as part of a wider Public Safety Master Plan for the capital. Officials frame the AI emergency response system as a way to sharpen how quickly calls get answered.
Home Ministry of Malaysia — Malaysia's Home Ministry will adopt AI police report drafting and processing for roughly 400,000 police reports a month nationwide. Home Minister Saifuddin Nasution Ismail expects AI first versions to ease administrative workload, though investigating officers must still check facts before AI-drafted reports go to deputy public prosecutors.
Ministry of Digital Transition and Administrative Reform (Morocco) — Morocco has named the AI in public administration it means to field: Idarati AI, an administrative helper, and a sovereign meeting assistant transcribing Arabic, French and Darija. The Ministry of Digital Transition and Administrative Reform says its Maroc IA 2030 roadmap has turned to delivery. Also shown: opening Jazari institutes, a Data Factory, a Software Forge and a sovereign language model covering Darija, Amazigh and Moroccan context, widening artificial intelligence tools for government services.
Public Procurement Service of Korea — Seven homegrown systems deliver AI in government procurement across South Korea's Public Procurement Service: procurement-law guidance, press release and remarks drafting, document automation, National Assembly task tracking, and workflow automation. A 26-person internal team built these AI-driven public sector automation tools since February 2026 via 'vibe coding', letting non-programmers specify requirements conversationally. Commissioner Baek Seung-bo unveiled the outcomes.
Märkischer Kreis — Since 11 September the Märkischer Kreis has picked up at any hour, passing callers to Emka, an AI telephone assistant for citizen enquiries. Emka steers people towards the right desk, paperwork or colleague, works 365 days a year and copes with over 30 tongues. Outside opening times, weekends and bank holidays included, the district's AI switchboard fields the call; by day the human counter answers first.
Ministry of Trade and Industry, Singapore / Enterprise Singapore — BizSG launches 30 September 2026, as Singapore's Ministry of Trade and Industry and Enterprise Singapore fold schemes into one BusinessSG roof. Its AI grant recommender reads a firm's goals to point it toward matching funds, an early case of AI-based business grant matching already used by roughly 700 firms in beta since October 2025. Low Yen Ling unveiled the plan at the SME Centre Conference.
State Customs Committee of Belarus — Belarus's State Customs Committee plans a customs AI application shifting from electronic to "intelligent" processing, applying machine vision to shipping and commercial paperwork attached to declarations. The system will read and cross-check these documents against the declaration, plus scan vehicle images to match each vehicle against its past crossings and similar vehicles, flagging discrepancies to officers. This border control automation targets document and vehicle verification, surfacing anomalies for human review.
Dubai Police, Digital Dubai — Dubai Police, with Digital Dubai, has switched on an AI shift management system for police workforce planning. The tool weighs the specialisations and skills of officers against operational requirements, recommends personnel distribution inside approved controls, and reaches departments and staff through the Smart Employee application. Digital Dubai means to offer the AI rostering tool to government entities across the emirate.
Independent National Electoral Commission — The Independent National Electoral Commission has rolled out an intelligent Results Management process for AI election results verification, flagging clashes between polling-unit tallies. Chairman Professor Joash Ojo Amupitan, at the Global Conference on Responsible AI and Elections in Ottawa, said machine reasoning should fortify, never supplant, voting bodies. Nigeria targets 2027.
Seoul Metropolitan Government — Seoul Metropolitan Government will spend 2.8 billion won (US$2.1 million) on an AI emergency response system for the city's waterways. Twenty-four audio sensors on bridge pillars pick out screams, impacts and splashes, fix a person within 10 metres and relay acoustic rescue alerts over private 5G in three seconds. A drone lifts off inside 60 seconds; Banpo Bridge takes an October trial before November's full launch. Fire brigade records count 1,272 suicide attempts on the Han River's 31 bridges during 2024.
Agency for Nature and Forests (Flanders) — Flanders' Agency for Nature and Forests says it will put cameras over protected land whose artificial intelligence reads the picture and picks out a rising plume of smoke. Validation runs through 2026 in North Limburg and other representative sites, with live trials this autumn and a staged roll-out from 2027 giving reserves AI wildfire detection. Minister Jo Brouns was blunt: every minute counts once a blaze starts, and early fire detection is the best chance of containing it.
South African Police Service — South African Police Service (SAPS) officers would wear body cameras capable of facial recognition under RFB 3286-2026, a tender issued through the State IT Agency (Sita). Also covered for three years: patrol-vehicle dashcams whose embedded AI must read number plates and spot blue lights, opened doors or crashes, plus supporting software. Face-matching is compulsory, so non-compliant bids are rejected before scoring. Submissions close on 29 September.
Agence Nationale de Sécurité d'État (ANSE) — Mali's ANSE runs mass SIM surveillance over roughly 25 million subscribers on three domestic networks; a linked consultant used Claude AI as the main build tool, Anthropic found. The telecom monitoring platform 'Lakana 360' harvests call logs, SMS, and voice data, spots speakers by voice, flags VPN use, and matches citizens against the biometric registry — no court approval needed.
Liberia Revenue Authority — The Liberia Revenue Authority is building an AI-powered contact centre for tax enquiries, aiming to give people answers on customs and duties without a manual lookup. Staff from the Taxpayer Services Division met partner Kwagei Group at a Margibi County retreat to shape the underlying knowledge base, overseen by the LRA Tax Knowledge Governance Committee. Complaints and service questions will also route through the new AI customer service system once live.
Ministère de la Justice du Sénégal — Senegal's Ministry of Justice launched two digital court platforms, e-Justice and Jokko ak Yoon. e-Justice standardises and traces court case processing; Jokko ak Yoon gives citizens SMS and audio access to the justice system in national languages. UNDP supported court digitisation built by the Direction de la Dématérialisation et de l'Automatisation numérique.
California Office of Data and Innovation — California's Office of Data and Innovation is piloting AskCA, an AI-powered digital assistant giving residents a single front door to state and local services, grouped by life events such as looking for work, launching a business, family support and disaster recovery, Governor Gavin Newsom said. Beta testers can now sign up; Anthropic's Claude underpins its knowledge base, with security help from the California Department of Technology, following early trials with 140-plus fire-recovery leaders and Altadena wildfire survivors. CalHR will first run it live, powering AI job-skills matching on CalCareers from 30 September.
City of Mountain View — The City of Mountain View, California, is piloting Wordly's AI translation at public meetings, letting residents follow in more than 60 languages; it reached a city council meeting on 8 September after debuting at an Environmental Planning Commission meeting a month earlier. The $58,367 pilot, including a $43,750 Wordly contract for 300 hours of live software translation, extends language access beyond the in-person Spanish and Chinese interpretation offered on request.
Iraqi National Security Service — Iraq's National Security Service has detailed 'Mayyiz', a cybersecurity-directorate platform for AI deepfake awareness that quizzes users on telling genuine images, video and audio from fakes and explains online fraud and disinformation methods. Billed by the service as the country's first specialised platform on AI and deepfake risks, the fake-spotting test is in preliminary trials via the government's 'Aman' platform, which hosts its data.
Municipalidad de Santiago de Surco — The Municipality of Santiago de Surco in Lima has introduced Norma, a generative AI municipal agent answering residents day and night on local taxes, procedures, services, activities and the cultural agenda, using only official information from the municipality's database. Residents reach the virtual agent via WhatsApp or the municipal website, and Surco is reportedly Peru's first municipality with an AI virtual agent.
Election Commission of Pakistan — Pakistan's Election Commission will launch an artificial intelligence wing for internal technical support, part of a broader AI in election administration drive, alongside a fully digital E-Office system due by December 2026. Sikandar Sultan Raja has set 15 October 2026 as the deadline for all wings to complete E-Office training — one step in the commission's electoral body automation. A new interactive public site was also approved.
Office for IT and eGovernment — Serbia's Office for IT and eGovernment will launch an AI assistant for passport appointments by late September 2026, automating identity card and passport bookings on the eGovernment portal. Director Mihailo Jovanovic told RTS users will ask, in natural language, for the nearest police station opening. Testing is underway, with plans to extend this conversational AI for government services to school and kindergarten enrolment.
Government of Sint Maarten — The Government of Sint Maarten is bringing an AI-supported HR system to its public sector, built on Microsoft Dynamics 365 to modernise how the civil service manages personnel records. Governor Ajamu Baly revealed the electronic human-resource management (e-HRM) initiative while opening the 2026-2027 Parliamentary Year, framing it as a move from scattered paper files toward one integrated database powered by data-driven human resource management tools.
Nigerian Army — The Nigerian Army is broadening data analytics, artificial intelligence and simulation technology into military AI decision support, for sharper battlefield judgement. Chief of Army Staff Lieutenant-General Waidi Shaibu billed simulation a critical command instrument behind predictive capabilities. Shaibu opened a Simulation Training Seminar at the Land Forces Simulation Centre Nigeria, Giri, Abuja.
Dane County — Dane County has put an AI chatbot for non-emergency calls, the Aurelian Virtual Assistant or AVA, on the 608-255-2345 line, fielding information requests, department referrals and routine questions since Aug. 18. A one-year contract worth $155,000 covers the software, 90% of that from grants and 10% from county funds. Officials stress that no human communicator or dispatcher is being replaced: the AI call handler exists to leave staff free for emergencies.
Ministry of the Interior and Safety — South Korea's Ministry of the Interior and Safety will grant promotion points, cash and lecture fees to officials wielding AI tools in the civil service. Beneath it sits the AI Government Lab, running since July, where staff build prototypes by talking to a coder; the ministry plans to lift its ceiling from 200 seats to 6,000. Government AI development logs 2,073 builders and 770 projects, 376 shared, chasing 20,000 AI Champions, near 2% of administrative and public-institution employees, by 2030.
Ministry of Higher Education and Scientific Research (Algeria) — Algeria will move to AI tax collection, Minister Kamel Baddari of the Ministry of Higher Education and Scientific Research told the opening inter-ministerial session. A second Algerian platform, spanning many sectors, should push the technology further into public services. Five committees now own delivery, and the automated tax administration work sits beneath a wider goal of 50,000 trained specialists by 2030.
Transport Bureau (DSAT) of Macao — Macao Transport Bureau (DSAT) has opened an AI Analysis Platform plus an online application system for AI vehicle type approval of brand and model filings. Dealers' paperwork is pre-checked before formal lodging, flagging mistakes and gaps. Built-in machine translation should speed AI-assisted approval work and shorten turnaround.
General Directorate of Identity and Foreigners Affairs, Dubai — Dubai's General Directorate of Identity and Foreigners Affairs launched 'Click and Travel', which lets officers clear a passenger's immigration transaction in roughly 12 to 15 seconds through biometric immigration clearance. The mobile immigration service combines live passenger records, automated pre-screening and electronic passport reading with facial recognition, and handled 31,373 passengers in a May-to-end-August pilot.
National Centre for Statistics and Information — Oman's National Centre for Statistics and Information released the Household Income and Expenditure Survey, applying AI in official statistics to generate indicators efficiently and sustainably. Calling the AI-produced survey a national, world-first achievement, the centre puts average monthly Omani household income at RO1,839.3 for 2025, up 18.5 per cent on 2018.
Tribunal Supremo de Justicia — Bolivia's Tribunal Supremo de Justicia is piloting AI assistants for judges and litigants together with a digital case file, court president Romer Saucedo confirmed. Paper case files still dominate the courts and, in Saucedo's account, invite bribes for sight of a file or a photocopy, whereas an online file would reach a citizen by mobile phone. AI in the courts arrives at a judiciary living on roughly 0.35% of the national budget and pressing for 5%.
French state (Ministry of Public Accounts) — The French state has pledged six million euros to Mistral AI across 2026 and 2027, buying AI for government administration across ministries. Les Échos broke the story on 2 September; the Ministry of Public Accounts confirmed a non-exclusive deal won by tender, with Mistral AI engineers posted into departments. Cybersecurity, fraud hunting and clearing bulk caseloads at the Ministry of Justice lead this public-sector AI push.
Ministry of Public Security — China's Ministry of Public Security has launched national anti-fraud AI, an AI anti-fraud app helping the public spot telecom and online scams, built independently by the Shanghai Public Security Bureau under its criminal investigation bureau. Users describe a suspicious situation and the app rates the risk, names the probable fraud tactic, gives prevention advice and points to similar cases, using AI agent technologies, multimodal models and large language models.
Dubai Police — Dubai Police has deployed 36 robots for robotic process automation in policing and administration, automation robots that take routine admin work off staff. The force also finished 104 strategic and operational technology projects last year, disclosed at Commander-in-Chief Lieutenant General Abdulla Khalifa Al Marri's annual inspection of the General Department of Artificial Intelligence.
Ministry of Finance, Nepal — Niraj Bhusal built an overnight AI disaster response portal for Nepal's Ministry of Finance after the Rasuwa-Bhote Koshi floods, merging satellite APIs, news reports and press releases with data from the Department of Hydrology and Meteorology, Nepal Police, the Nepali Army and the National Disaster Risk Reduction and Management Authority to track deaths and missing persons for relief coordination. An AI flood-response platform launched in Nepali, then English.
Government of Punjab — Punjab chief minister Maryam Nawaz Sharif opened the Provincial Intelligence Fusion and Threat Assessment Centre in Lahore and e-launched AI Hub Punjab, which she calls Pakistan's first provincial AI hub. The centre combines intelligence from multiple agencies and alerts district coordination committees digitally, catching problems before becoming major incidents.
National Database and Registration Authority (NADRA) — More than three million Pakistani retirees now confirm pension eligibility through AI face verification pensioners technology instead of fingerprint scans, under a newly activated National Pensioners Verification System run by the National Database and Registration Authority (NADRA). Coverage spans federal, provincial, military and autonomous-institution pension schemes, targeting older and disabled claimants whose worn prints routinely defeated fingerprint readers, with Pakistan's Prime Minister launching the system on 20 August 2026. Each successful scan triggers an automatic digital Proof-of-Life Certificate, logged straight into the Pension Controller Register System without manual paperwork. This marks a shift toward facial biometrics for pension verification across Pakistan's public retirement system.
Supreme Court of the Russian Federation — Russian courts across several regions are piloting four large language models trained using judicial practice, Supreme Court Chairman Igor Krasnov says. This test of AI in courts draws on legal positions and practice-forming rulings. The Supreme Court also endorsed a court AI integration concept and created a Centre for Judicial Competences in AI.
Ministry of National Defence of Lithuania — Lithuania's Ministry of National Defence issued a roughly €5 million procurement for AI airspace surveillance, merging military sensors, radars and other agencies' feeds live. Defence Minister Robertas Kaunas expects stronger air-threat warning and response by state institutions from this AI drone detection network. Piloting begins early 2027, full sensor coverage follows within 2027, and NATO allies' platforms connect by 2028.
Swiss Parliamentary Services — Swiss federal parliamentarians are getting PIA, an in-house AI assistant, for the autumn session. Users include council members and their personal staff, Parliamentary Services employees and parliamentary group staff; for now it works best on committee documents rather than internet research. Swisscom runs the project, keeping operation and data storage in Switzerland, with up to CHF 150,000 budgeted for a roughly year-long pilot.
Poder Judicial del Estado de México — The State of Mexico's judiciary has run 1,401 court rulings through an AI tool producing plain-language rulings for citizens, Governor Delfina Gómez told its first 2025–2026 annual report in Texcoco. It belongs to the court's 'internet of rulings', alongside an electronic court platform the governor says has served over 170,000 users.
UAE Cabinet — UAE Cabinet has approved a Cabinet AI Advisor system of 32 specialised AI policy advisers working with ministers around the clock. These digital advisers review policy and legislation, gauge impact, benchmark global best practice and recommend; Sheikh Mohammed bin Rashid Al Maktoum approved it at the season's first Cabinet meeting, at Qasr Al Watan in Abu Dhabi.
Dubai Land Department — Dubai Land Department's Initial Registration platform brings AI property registration to developers, reading Emirates IDs, passports and sales contracts, extracting the relevant data and auto-filling fields. Project and transaction registration and escrow account management now share one system; senior manager Mustafa Al Rifai says data entry drops up to 80 per cent and 30-minute registrations take under five.
Government of Jordan — Jordan's government has introduced Farah, an AI government spokesperson created as a virtual character, to inform citizens and others, explain decisions and policies and speed official news, says Government Communications Minister Mohammad Al-Momani. Farah has an Instagram account, and Al-Momani later announced an English-language version of the virtual spokesperson.
Gobierno de la Provincia de Santa Fe — Santa Fe's provincial government opened Penitentiary Unit No. 9 in Recreo, 891-place prison with AI prison surveillance and facial recognition to spot wanted people, visitors and potential risks. Voice identification recognises detainees and approved callers and ends any call automatically if someone else speaks. Governor Maximiliano Pullaro likened it to Lince, Rosario's AI camera system, and called it the country's most technologically secure prison.
Panamá Conecta / Centro de Atención Ciudadana 311 — Panamá Conecta, the Panamanian government's single public-services portal, now has an AI government chatbot for fast reports to the 311 Citizen Attention Centre. The 311 platform also offers live chat with a contact-centre agent.
Politie Gent — Politie Gent's new AI chatbot in policing answers residents 24 hours a day on the force's site, following months of trials. It greets visitors as the Ghent force's assistant, handling routine queries so officers, per the force, gain more room for genuine face-to-face contact — part of a wider shift toward AI in law enforcement.
Ministry of Industry and Technology (Türkiye) — Türkiye will compel AI training for civil servants across the whole payroll, a drive steered by the Industry and Technology Ministry. Lessons run from generative use and privacy to ethics and synthetic video — artificial intelligence literacy in the public service. Ministers want 30 trials across at least 10 bodies in two stages inside 12 months, five or more citizen-facing offerings on the e-Government portal, and three-year roadmaps in 12 weeks.
Uganda Police Force — An AI chatbot for police services is planned by Uganda Police Force, speeding citizens' access to emergency guidance and general information. Dr Aisha Kayongo unveiled the plan at the Wakiso District ICT facility, framing it as a simpler route for people confused about crisis response. This AI-powered policing assistant supplements 999 rather than replacing it.
State of Nebraska — Nebraska has chosen Tyler Technologies' Resident AI Assistant for statewide use after a May–August pilot at the Department of Motor Vehicles. During the pilot, the state says, it answered over 88,000 questions from 38,000-plus users, made more than 3,400 referrals to state online services and cut call volume by 20 per cent in some areas, with nearly a fifth of interactions after hours. It launches on Nebraska.gov as Nebraska Navigator before reaching participating agencies, and the DMV is widening its use of Tyler's document automation for motor carrier operations; deputy director Mitch Greenwall said the pilot showed AI improves resident experience while easing pressure on agency resources.
Comune di Pieve Ligure — Comune di Pieve Ligure has deployed 43 artificial intelligence systems for AI in local government, cross-checking municipal paperwork against regulations and flagging discrepancies to staff. Trialled across the general secretariat, local police and technical office, the tool won over staff, so the council funded a wider rollout using the ordinary budget. Small municipalities like this one, population roughly 2,500, rely on AI for municipal administration where single employees often run entire offices alone.
US Dept of Defense — The US Department of Defense activated ChatGPT Mil, OpenAI's model, on its GenAI.mil platform, accredited for controlled unclassified information at Impact Level 5, alongside Google Gemini and Grok for Government. The platform serves document-heavy planning, policy, logistics and administration across a workforce of more than 3 million. The Pentagon reports GenAI.mil already has over 1.7 million unique users, and the Navy and Marine Corps require it enterprise-wide.
Toronto Police Service — Toronto Police Service has deployed AI-powered cameras that read licence plates plus vehicle makes and models, which the service rates highly effective against crime. The system logs where and when vehicles are spotted and raises alerts officers can act on. Privacy researchers warn of possible misuse and mass surveillance while automatic plate recognition spreads through Canadian forces.
Gobierno del Estado de Tamaulipas — The Tamaulipas state government will turn AVi-T into an AI citizen assistant answering questions on WhatsApp and a web portal about vehicle-tax payments, birth certificates and the right office for each procedure, from vetted institutional information. Edwin Tuexi Amaro, who heads the new Secretariat of Innovation and Digital Intelligence, says the virtual assistant goes live right after the governor's launch within weeks, making citizen-service staff more efficient without cutting headcount.
Armed Forces of Montenegro — Digital Adjutant, a new AI system for military command staff, was built in-house by Montenegro's Armed Forces for information analysis, legal drafting and document preparation. Minister Dragan Krapović said military IT specialists created it to ease administrative workload and free personnel for harder duties. Defence officials called it a domestic AI capability for national defence, keeping sensitive data under full digital sovereignty instead of foreign systems.
Albanian government (Smart City project) — Twenty schools in Tirana, Albania now run AI school security cameras under a Smart City initiative from Prime Minister Edi Rama. Rather than merely storing footage, the system spots and alerts staff to suspicious activity as it happens. The capital's 20 sites mark the programme's opening stage, ahead of any wider rollout of smart surveillance cameras in schools.
Fuerzas Armadas de Paraguay — Paraguay's armed forces and the National Emergency Secretariat have launched Operation Arapochy to prepare for El Niño flooding on the Paraguay River, using AI flood planning from an in-house AI agent. It blends live weather and hydrological feeds from open sources such as NOAA and Windy with historical climate models, and military simulations put the flood peak between November and December, so preventive evacuations can start in October. Planning covers up to 100,000 affected families.
Secretaría de Planificación y Programación de la Presidencia (SEGEPLAN) — Guatemala's SEGEPLAN now runs Copiloto SNP, offline AI in public sector planning that verifies and enters National Planning System data. No internet means no token charges, so officials handle sensitive files directly. It also unifies quantitative and qualitative report variables, trimming duplicated effort as agencies adopt AI for government administration.
US CIA — The US CIA's Open Source Enterprise used GPT-4.1 to help compile open-source intelligence reports on the 2026 FIFA World Cup, sharpening article selection, translating titles and summarising host-country coverage. Drawn from unclassified, publicly sourced material and checked by CIA analysts at every stage, the reports went to local law-enforcement partners via fusion centres such as the Washington State Fusion Center.
US Dept of Veterans Affairs — The US Department of Veterans Affairs is beta-testing an LLM-driven help-desk chatbot on its 'Contact Us' page, letting veterans ask conversationally about benefits and services. It steers users to resources but offers no route to a person, is unsuitable for emergencies or medical questions and gathers no personally identifiable information, the VA says.
Soka City (Saitama) — Soka City in Saitama has deployed FIXER's Sovereign GaiXer, on-premise generative AI for drafting internal rules, preparing meeting minutes and building plans from confidential documents. The city says it is Japan's first municipality to adopt it, keeping sensitive data on site, away from public cloud services.
Ministry of Science and ICT — South Korea's Ministry of Science and ICT has picked SK Telecom, KT and Kakao from six applicants to lead consortia for citizen AI services: a general-purpose chatbot and AI agents for complex public administrative procedures. Agreements follow in September, nationwide launch targets end-2026 after beta tests, and the ministry will supply the three with 512 Nvidia B200 GPUs this year.
City of Calgary — The City of Calgary has used an AI platform to score over 17,000 intersections lacking signals or flashing beacons from zero to 100 for pedestrian risk, publishing the 50 riskiest. All scored 93 or more, led by Martindale Boulevard and Martindale Gate N.E. (99.02). Staff recommend kerb extensions, rapid flashing beacons and similar fixes at 26 of them, costing roughly $6 million, with design in 2027 if council approves funding in November's budget talks. Ward 5 councillor Raj Dhaliwal backs data-driven decisions but wants the model's reasoning explained; last year 15 people died on Calgary streets.
Dirección General de Aduanas (DGA) — The Dominican Republic's General Directorate of Customs (DGA) will add AI X-ray cargo inspection, with an advanced model reading non-intrusive scan images together to flag risk patterns, possible inconsistencies and items needing specialist review. Director general Nelson Arroyo told an Association of Industries of the Dominican Republic meeting that headquarters also wants live oversight of X-ray operators at ports and AI in its risk-management engine for sharper cargo profiling.
National Police Service, Ministry of Interior — Kenya's Interior Ministry will deploy AI-powered police deployment technology to direct officer placement nationwide, per Kipchumba Murkomen. Data analytics will drive the AI-driven officer staffing, linking digital occurrence books and command centres across six cities, with CCTV cameras at 1,200 police stations. Completion is targeted for December 2026.
Ministry of Defence of Ukraine — Ukraine's Ministry of Defence has opened its military AI assistant to everyone on the DELTA combat system, trimming web-widget help tickets by roughly 27% and shortening replies by a factor of 30. Hosted inside DELTA's protected environment, this combat-system AI helpdesk tackles routine enquiries on-screen and fields around one in four widget messages.
National Board of Revenue, Bangladesh — Bangladesh's National Board of Revenue (NBR) is developing Korshathi, an AI chatbot for tax filing to demystify income tax, VAT and customs duties. A World Bank-funded revenue mobilisation scheme funds the effort, letting the tool render dense rules into everyday language while pulling citizen records through National Identity Card numbers. Officials describe it as a digital tax companion, streamlining access to state services.
Yunlin County Government — Yunlin County Government, Taiwan, now runs an AI for civil servants named Document Brain, using Foxconn's CityGPT platform. Roughly 20,000 records over seven months taught this AI agent for government document drafting to verify facts and surface knowledge; people still approve every result.
Prefeitura de Florianópolis — Florianópolis city council has launched 'Tia', a WhatsApp generative-AI assistant for over 600 municipal services, from tax and health guidance to business licensing and company registration. Softplan built the round-the-clock, multilingual AI government chatbot, which reads text and images and passes users to staff; the council calls it the first generative-AI public-service assistant in a Brazilian state capital.
Aduana Nacional de Bolivia — Bolivia's Aduana Nacional has launched AranBUS, AI customs classification running on natural language processing to speed dispatches and cut misclassification of goods. A basic level answers quick, exact queries; an advanced level reads synonyms and dense technical context for operators and the wider public. The AI tariff search came bundled with RecaudoMETRO, posting key tax collection openly every seven minutes, and the Mi Aduana app for entrepreneurs, operators and travellers. Executive president Alberto Soto de la Vía called it a customs service that is modern, efficient and fully transparent.
Digital Policy Office — Hong Kong's Digital Policy Office (DPO) is building an AI government hotline, launching four AI Efficacy Enhancement Projects for 1823 by end of 2026. LLMs and speech recognition power a Call Assistant drafting case summaries, an Email Extraction tool cutting data entry by 30%, priority sorting of department replies, and a district-level Dashboard. AI Agents get piloted across departments in Q4, widening automation into public-sector case handling.
MeteoSwiss — Switzerland's MeteoSwiss has activated an AI thunderstorm forecasting system after a four-year build. It rescans nationwide radar and lightning readings every five minutes, flagging storm-hazard odds within the following hour. This deep learning storm prediction engine trims false alarms and lengthens warning lead times versus MeteoSwiss's prior setup.
French government — Mayotte will gain new AI coastal surveillance under a French plan Sébastien Lecornu confirmed, aimed at curbing seaborne illegal immigration. Radar upgrades, a fixed lookout base at M'tsamboro, long-range drones and trial balloons for constant altitude watch make up the toolkit. Officials called it a move toward stronger "militarisation" defending the island's sovereignty, part of wider AI-assisted border surveillance efforts.
Customs Administration, Ministry of Finance — Under a five-year Smart Customs Navigation Plan unveiled 26 August 2026, Taiwan's Customs Administration is rolling out AI customs services. A knowledge base merges with agentic AI to answer questions, while a separate function scans photos or written product details to recommend cargo tariff codes. Staff trials come first, ahead of a customs AI assistant dashboard reaching officers by 2027.
Tanzania Police Force — Tanzania Police Force runs AI cybercrime investigation tools to trace online offences, says Joshua Mwangasa. Syndicates faking voices with AI drove the shift toward automated cybercrime tracking. Five hubs opened in Mwanza, Mbeya, Arusha, Dodoma and Zanzibar, broadening coverage past the capital.
Local governments (Jiangsu, Beijing, Shanghai) — Local governments across China have deployed grassroots AI tools: in Suqian, Jiangsu, community-group AI assistants answer social security and school enrolment questions and route feedback. In Beijing's Miyun district an official used AI-assisted coding to build a flood-control monitoring app, and a Shanghai subdistrict launched a community-service agent with voice Q&A and public-service maps. No adoption figures were given.
Odašiljači i veze — Odašiljači i veze has widened OIV Fire Detect AI, Croatia's AI wildfire detection grid, by 50 watch cameras across 25 fresh sites, favouring protected land such as national parks and nature parks; the estate now stands at 143 locations and 282 cameras. Advanced algorithms pick out smoke by daylight and flame after dark, engineered for unbroken cover 24 hours a day, seven days a week, year-round, with every county fire monitoring centre able to pull pictures and readings. Procurement fell to the Ministry of Environmental Protection and Green Transition; the forest-fire detection network was paid for through the Competitiveness and Cohesion Programme 2021-2027.
AADE (Independent Authority for Public Revenue) — Greece's Independent Authority for Public Revenue (AADE) has activated dedicated units for AI tax audits, running risk analysis and automated cross-checks. They are meant to improve how the authority spots tax risk and audits.
KPRM, Senate and Prokuratoria Generalna — Testing of an AI assistant for public administration is underway at three Polish bodies: the Chancellery of the Prime Minister, the Senate Chancellery and the Prokuratoria Generalna (General Prosecutor's Office). Deputy Digital Affairs Minister Dariusz Standerski described three planned uses: answering general questions on central government, a legal function that cites regulations plus relevant court decisions, and automatically anonymising documents. Built on the PLLuM language model, this AI tool for government officials should see two or three functions rolled out into central office systems by end-2026.
Inland Revenue Department (IRD) — Sri Lanka's Inland Revenue Department has rolled out RAMIS 3.5, letting AI tax audits target returns by risk score instead of officer judgement. The Revenue Administration Management Information System now runs as an API-driven, real-time compliance engine, narrowing chances for corrupt interference and replacing risk-based audit selection once left to manual discretion. It is IMF-backed under the Medium-Term Revenue Strategy and links to compulsory real-time e-invoicing for VAT-registered firms.
Ministry of Communications and Information Technology (MCIT) and Cairo Court of Appeal — Cairo Court of Appeal will deploy AI for judges under a three-year protocol with Egypt's Ministry of Communications and Information Technology. Planned tools include intelligent search, drafting memoranda and working papers, summarisation, and cross-checking rulings, alongside a digital archive, public site and app, and staff training. This judicial AI deployment targets faster case turnaround, not replacing judges' verdicts.
Digital Agency (Government of Japan) — Japan's Digital Agency launched Genai, in-house generative AI for civil servants, to about 180,000 officials for translation, writing, meeting transcription, public-comment classification and draft replies to parliamentary questions. Fact-checking and final policy calls remain with staff; broader rollout is aimed at fiscal 2027.
UK Home Office — The UK Home Office now uses AI to triage and route police 101 non-emergency calls, aiming to cut waiting times. The step belongs to wider work to reduce pressure on non-emergency phone lines.
Conselho Nacional de Justiça / Tribunal Regional Federal da 3ª Região — Brazil's National Council of Justice has added LIA3R, generative AI for courts built by judges and staff of the Federal Regional Court of the 3rd Region, to its Conecta programme, opening it to every court nationwide. The judge-built AI condenses long filings, handles specialised legal research and writes decisions, dispatches and case summaries, with about 1.55 million documents summarised across 106,000 cases and roughly 100,000 interactions a month, the court says. It launched at the sixth FestLabs in Cuiabá.
Government of Armenia — Armenia's 2026-2031 government programme commits the force to a 'Digital Police', with an AI-based 'AI Police' unit operating online to keep citizens safe, run a cyber laboratory, score risks, profile people and offer services 24/7. AI policing extends to criminological analysis, with fresh tooling to read offence patterns, where they cluster and what threats follow, plus new methods for projecting crime trends, spotting danger sooner, sharpening prevention and grounding decisions in evidence. Further steps target hostile foreign information manipulation, falsehoods and fabricated material, including deepfakes and synthetic AI content.
Shanghai Municipality — Shanghai switched on 100 AI-powered tourist service terminals across the city, each fronted by an AI agent named Hu Xiaoyou that handles navigation, real-time translation, attraction recommendations, ticket booking and made-to-order itineraries. The AI tourist service terminals draw on more than 84,000 entries covering cultural activities, attractions and sports events, carry 14,000 booking and shopping links, work in ten languages — Chinese, English, Japanese, Korean, French, German, Spanish, Thai, Russian and Arabic — and take payment in more than 120 currencies. Machines stand in major hotels, tourism information centres, event venues, business districts and scenic spots, and volunteers will also be stationed with the AI visitor assistant to help visitors with additional needs.
Transport Department — The Transport Department said its Easy Licence Application Approval project will go live by the end of 2026 and automatically recognise, cross-check and verify the documents behind online vehicle licence renewals, pushing automatic approvals past 90 per cent and cutting processing from up to ten working days to one. The same AI vehicle licensing work is to cover the Northbound Travel for Hong Kong Vehicles and Southbound Travel for Guangdong Vehicles schemes, where the department is aiming at over 50 per cent of cross-boundary driving permit applications approved automatically and manual processing times about 20 per cent shorter. Automated licence approval is the largest of a set of AI projects the department set out that day.
City of Kigali — Kigali now machine-sorts rubbish: the City's pilot runs an AI waste sorting system that scans waste, flagging it hazardous, plastic, e-waste, paper, organic or otherwise. Built on the Smart Eco-Platform, it links an app to IoT smart bins and an admin portal. COPED Ltd's Buregeya tested the AI-powered waste sorting system at Kigali's 29th International Trade Fair, now expanding to top refuse sites.
Conselho Nacional de Justiça (CNJ) — Brazil's CNJ spread generative AI for courts nationwide via Conecta: three Mato Grosso state court tools. Within this judicial AI, LexIA analyses processes and drafts, Hannah checks appeal admissibility, OMNIA clears case backlogs with real-time recommendations.
UK West Yorkshire Police — West Yorkshire Police deployed a live facial recognition van in Bradford city centre, arresting four people wanted for crimes, finding a missing person and flagging 21 people under Sexual Harm Prevention Orders for compliance checks. It was the force's seventeenth such Bradford visit since February 2026, with 29 arrests in all.
Egyptian Tax Authority — The Egyptian Tax Authority's mobile app for real-estate transaction tax uses AI identity verification against a national ID or passport, so citizens register property deals and pay online. Beyond the AI ID check, the app gives indicative prices from actual market values and serves Egyptians abroad without a tax-office visit.
California Department of Technology — Poppy, the California Department of Technology's generative AI tool, routes work to Claude, Gemini and ChatGPT with data kept in state systems, and now serves about 1,000 staff in 24 departments. Built on existing budget after a 2025 pilot (2,700-plus people, about 70 departments), it charges departments $7,500–$35,000 to join plus per-user fees. Lucy Andrews of the Department of Water Resources finds the tasks it absorbs often dull and slow. Users remain well below one per cent of state staff.
UAE Federal Government — The UAE federal government's new National Agentic AI Project moves services onto agentic AI in government, keeping human oversight, with early workflows in technical support, customer service, tax auditing and procurement. Over 50 federal entities have joined, aiming to run half of federal services on autonomous AI agents within two years.
Alcaldía Municipal del Distrito Central — Tegucigalpa's city hall has launched MDC en tus Manos, a WhatsApp assistant it bills as Honduras's first municipal AI agent. Residents photograph and locate damaged signage, blocked drains, water leaks or potholes; the agent classifies and routes each report to the right department in under a minute and answers refuse-collection and water-distribution schedule queries. Built by the Alcaldía Municipal del Distrito Central's new AMDC+IA unit with the Universidad Tecnológica de Honduras' AI laboratories, it aims to reinvent bureaucracy, Mayor Juan Diego Zelaya says.
Digital Policy Office — Hong Kong's Digital Policy Office has run 30 AI 'Efficacy Enhancement Projects' across 13 government departments since mid-2026, spanning complaint handling, environmental protection and hygiene, leisure and sports facilities, and transport and food-business licences. It credits AI in government services with up to 90% shorter waiting or processing times in some cases and some internal workflows cut from days to hours by the AI projects; no independent evaluation exists.
Ministry of Railways / CRIS — India's Ministry of Railways, via its technology arm CRIS, has begun rolling out AI railway reservations to replace the 40-year-old Passenger Reservation System, including forecasts of whether waitlisted tickets will confirm. Indian Railways says the waitlist prediction, introduced earlier in 2026, lifted accuracy from 53% to 94%, unverified independently; about 88% of tickets are already booked digitally.
Andhra Pradesh Commercial Taxes Dept — Andhra Pradesh's Commercial Taxes Department has put forward an AI/ML-driven GST model, AI tax evasion detection that also speeds refunds. It is pitched as a way to strengthen the state's GST administration.
Prefeitura de Rio Verde (Goiás) — Rio Verde's city government in Goiás runs an AI municipal chatbot from Gove covering 244 services across 24 departments through 26 specialised AI agents. Since January 2026 it has handled nearly 65,000 interactions, settling about 40% of requests without staff, the city says; no independent evaluation is published.
Ministerio de la Presidencia, Justicia y Relaciones con las Cortes — Spain's Ministry of the Presidency, Justice and Relations with the Courts reports saving over one million hours through AI in courts and robotic process automation, which completed 350,000 administrative tasks and handled procedures worth more than €700 million. Court automation tools span automatic transcription, data anonymisation, summarisation, document classification, case-file information extraction and AI-powered translation for citizen services.
UK British Transport Police — British Transport Police has taken its live facial recognition trial into London Underground stations, deploying NEC's NeoFace M40 at Victoria after starting at London Bridge in February. Concourse cameras check passengers against a police watchlist for high-harm offenders, especially people sought for sexual and other serious crimes. BTP says non-matches are deleted immediately; the pilot, first set at six months, now runs to November 2026.
Ministry of Digital — Malaysia's Ministry of Digital added agentic AI for public services to the MyGOV Malaysia app, letting citizens describe needs in everyday language and get step-by-step help, such as looking up and then settling a police summons. The app had over 2.9 million users by July 2026 across 52 services from 19 agencies; the AI arrives in phases with attention to privacy and human control.
Government of Rwanda — The Government of Rwanda has started a nationwide satellite-data programme with Planet Labs through the Rwanda Space Agency, billed as Planet's first in Africa, serving agencies, public universities and selected startups. The near-daily imagery and data will aid disaster preparedness, urban planning, forest-health tracking and agriculture and food-security monitoring.
City of Cape Town — The City of Cape Town has unveiled Maya, an AI investment guide for firms eyeing the city, launched by Mayoral Committee Member for Economic Growth Alderman James Vos under the City's Cape Town. Built for Business campaign. Open 24 hours a day in eight languages including Afrikaans and isiXhosa, the AI guide takes users through municipal processes, investment openings and business support programmes and links them to relevant City officials.
Ministerio de Planificación del Desarrollo y Medio Ambiente — Bolivia's Ministry of Planning, Development and Environment launched BolivIA, state-built AI for legal research that lets public servants query laws, decrees and resolutions, citing the exact instruments. Minister Fernando Romero Pinto says research drops from hours to seconds, using only current regulation, in what the ministry calls step one of a wider state AI strategy.
US Army Human Resources Command — The US Army Human Resources Command rolled out Salesforce's Agentforce Public Sector platform for round-the-clock HR support to about 9.2 million military families, veterans, soldiers and civilian employees, via HRC's Digital Front Door personnel-records and HR-services platform. HRC says it is the first US military body on the platform since its Impact Level 5 authorisation, letting AI agents handle controlled unclassified information.
New Orleans — Orleans Parish Communication District — The Orleans Parish Communication District has put Carbyne's AI on live 911 calls to triage incidents and screen out duplicates, notably from rush-hour crashes, making New Orleans the first major US city to disclose one. Carbyne says the AI triaged over 3,500 events in 90 days, cut redundant calls over 30% and sped responses by up to 40 seconds, figures not independently evaluated. It learned from real 911 recordings, local street names and speech patterns.
US Dept of Defense — The US Department of Defense has authorised Salesforce's Agentforce 360 at Impact Level 5, clearing the agentic AI system to handle controlled unclassified information and some national security data. Offered via Salesforce's Missionforce National Security suite, it will support command insights, onboarding, logistics and administrative workflows on sensitive unclassified missions, the company says.
OPDC, NSTDA and DGA — Thailand's Office of the Public Sector Development Commission, with the National Science and Technology Development Agency and the Digital Government Development Agency, piloted an agentic AI chatbot for government on Pathumma, a Thai-developed large language model. It links the Rights and Liberties Protection Department, the Food and Drug Administration, the Department of Provincial Administration, the Revenue Department and the Department of Lands, so one citizen query spanning several bodies gets a single sourced answer.
Government of the State of San Luis Potosí — San Luis Potosí's government has put four AI robot dogs into public-security service, the first Mexican state to do so. Governor Ricardo Gallardo calls them technological first responders that patrol, scout terrain and temperature, go where officers would be at risk and stream live images to command centres. They debut at the Feria Nacional Potosina; 16 are earmarked for the Civil Guard, and five humanoid robots ordered from China have not yet arrived.
Ministerio de Defensa Nacional — Flagging risks ahead of signature is now the job of Helios, an AI corruption detection in public contracts system rolled out by Colombia's Ministry of Defense. Built to scan millions of SECOP I and SECOP II procurement entries, it hunts for warning patterns: bids with only one competitor, suspiciously compressed award windows, clusters of business going to the same supplier, and companies carrying past sanctions. So far the AI procurement fraud detection tool sits in an early pilot stage, restricted to internal use by ministry staff.
Protecció Civil — Protecció Civil wrapped up the wildfire-prevention drive with an AI-generated image showing the Madriu valley ablaze. Nothing in the scene is real; the body released the fake picture so Andorra grasps how close an actual blaze could be.
Ministry of Economy and Finance (General Department of Digital Economy) — Plain-language queries from Cambodia's finance staff now return instant tables, charts and summaries from 600-plus datasets on the Data EF Portal, via Data EF AI Agent, a government data query AI agent. The General Department of Digital Economy built this AI-assisted dataset search tool, cutting wait times to 10-15 seconds in trials, live per OECD since August 2026.
World Food Programme — The UN World Food Programme (WFP) uses AI food aid targeting in Somalia through its HungerMap Live platform, forecasting hunger hotspots and pre-positioning resources before conditions worsen. In Burhakaba district, this predictive humanitarian AI approach enabled WFP to deliver food assistance for 48,000 people and nutrition support for 3,000 women and children. HungerMap Live 2.0, launched April 2026, added forecasting features guiding priority decisions for communities.
Public Prosecution (Qatar) — Qatar's Public Prosecution and the Ministry of Interior have launched AI self-service kiosks letting inmates of penal and correctional institutions lodge requests by speaking in their preferred language. Requests reach the Execution of Judgments Prosecution, which reviews, decides and replies through the same AI system, and inmates can view case reports and follow each step.
Oficina Gubernamental de Tecnologías de la Información y Comunicación (OGTIC) — Taína, an AI assistant for public services from the Dominican Republic's Government Office of ICT (OGTIC), will handle passport renewals via WhatsApp, Facebook and Instagram. It flags expiring documents, fills forms, books appointments and takes payments, with human oversight, audits and fraud checks built into this AI-powered virtual government assistant. Five digitalised services launch in 2026, expanding afterwards, OGTIC's Edgar Batista said.
Dept of Public Service & Administration — South Africa's Department of Public Service and Administration has launched the 'Batho Pele' AI government chatbot with Meta and Juicetel, offering citizens and public servants instant answers on policies, circulars and legislation. Unveiled by Minister Mzamo Buthelezi on 31 July 2026, the policy assistant will grow to answer in nine official languages and backs human judgement rather than replacing it.
Government of Honduras (Sistema Nacional de Emergencias 911) — AI facial recognition emergency system upgrades are coming to Honduras's Sistema Nacional de Emergencias 911 under President Nasry Asfura, adding licence plate identification and vehicle data to sharpen nationwide response speed. Marking the service's tenth year, officials pointed to over 6,900 recovered cameras and faster reaction times, folding AI-driven public-safety surveillance into a decade-old network.
Federal Supreme Court of Ethiopia — More than 35 courtrooms in Ethiopia's federal high courts and first-instance benches now run smart-court upgrades, after the Federal Supreme Court switched on extra units backed by the European Union. AI-powered voice-to-text transcription in the judiciary, or speech-to-text conversion in court proceedings, turns testimony straight into text, replacing headphone-and-typing work by staff. President Tewodros Mehret tied this to faster, accessible justice.
National Directorate of Industrial Property (DNPI) — AI trademark examination has arrived at Uruguay's National Directorate of Industrial Property. Its in-house assistant scans submitted files—images, attorney authorisations, proof of payment—spots problems, and drafts notes that staff then confirm under this AI-assisted trademark registration approach.
Abu Dhabi Judicial Department — The Abu Dhabi Judicial Department launched an AI judicial platform that automates case-file analysis, quickly finds and compares legal precedents and helps draft legal memoranda and judicial documents. Judges retain full authority with the AI as legal assistant, in what officials call the world's first fully integrated AI-powered judicial platform.
National Tax Agency — Every tax office in Japan now runs predictive AI tax audits under the National Tax Agency (NTA), scoring filers for underreporting risk, spotting bogus refund bids and tracing corporate evasion schemes. Staff keep final say on whether scrutiny becomes a formal audit, while models pull from returns, financial statements, statutory records and audit history to rank cases. KSK2, a newer platform launched September 2026, broadens the dataset behind this tax risk-assessment AI.
Federal Board of Revenue (FBR) — Pakistan's Federal Board of Revenue has started building an AI-based 'faceless' tax system to spot anomalies and run AI tax audits without taxpayers and officials meeting. The faceless system is framed as part of reforming the revenue authority.
Ministry of Popular Power for Science and Technology (Mincyt) — AI damage assessment earthquake response now runs at Mincyt, Venezuela's Ministry of Science and Technology, which built the system after the double earthquake of June 24. Drone and satellite feeds train the model to spot, classify and flag structural harm, plotting results onto colour-coded maps graded from critical red down to minor yellow. Output reaches the armed forces, civil protection teams and the presidential infrastructure commission, giving relief coordination a shared, geo-tagged picture of where damage is worst — a form of AI-based infrastructure damage categorisation applied directly to disaster response.
Biosecurity New Zealand / Ministry for Primary Industries — Andrew Hoggard confirmed a four-month trial of generative AI biosecurity import rules drafting, run by Biosecurity New Zealand within the Government. The system drafts import health standards, freeing staff from paperwork while specialists keep final say. Ministers frame this AI-assisted biosecurity regulation as modernisation, not automated judgement.
e-Government Authority (e-GA) — Msaada AI, Muhtasari AI and Tafsiri AI — three tools from Tanzania's e-Government Authority and its Research, Innovation and Development Centre — bring AI-powered government services to queries, summaries and translation. Citizen feedback drives refinements as the Tourism Board website and President's Office e-Mrejesho tool already run this AI chatbot for public services daily.
Tesorería General de la República — Chile's Tesorería General de la República put an AI voice agent built with CGS Nexus on tax-query calls during Operación Renta, trained on 50-plus enquiry types. It served over 43,000 people and settled about 60% of calls alone, passing complex cases to staff, the Treasury says; no independent evaluation is published.
Minsk City Executive Committee — Minsk will add AI contact centre agents to the 115.bel housing hotline before the year is out, said Maksim Bardovich, who directs the Minsk City Executive Committee's Information Technology Centre. The agent-bots should shorten how long a request takes to handle and an answer to draft, sparing residents a queue. The hotline logged over 850,000 reports in the first half of 2026, routing them to city services.
Ministry of Labour, Employment, Veteran and Social Affairs — Serbia's Ministry of Labour, Employment, Veteran and Social Affairs put its first AI assistant for social protection queries, Radoslav, live on the eGovernment portal on 23 July 2026, running 24/7 without collecting personal data. The tool answers, conversationally, questions on qualifying for child support, family allowances and other welfare payments, plus application steps. Running on domestic infrastructure, including the Kragujevac supercomputer, the AI-powered social welfare chatbot clarifies rules rather than ruling on claims, leaving final decisions to human caseworkers.
General Department of Digital Economy, Ministry of Economy and Finance — Cambodia's General Department of Digital Economy has deployed an AI chatbot for e-invoicing on Telegram, an AI-powered invoicing helper on Telegram covering the Cambodian Electronic Invoicing System (CEIS). Called CamInvoice Chatbot, it sits under the Ministry of Economy and Finance to ease CEIS support.
Stadt Rüsselsheim am Main — Rüsselsheim am Main will introduce AI minute-taking from notes or audio, a semantic search and dialogue tool spanning archives, meeting papers and project documents, and an email assistant that sorts and ranks messages, part-drafts replies and condenses long threads. Hesse's state secretary for digital affairs handed the mayor 309,000 euros for the project under the Starke Heimat Hessen programme. The three tools will run on a German-based sovereign public-administration platform and later be reusable by other municipalities.
Ministry of Digital Transformation of Ukraine — Ukraine's Ministry of Digital Transformation, with Ukrainian developer Kitsoft, has added a voice-based government AI assistant feature to Diia.AI. Citizens can now speak queries and hear spoken replies instead of typing, and can switch between voice and text mid-conversation without restarting. The rollout extends this voice-enabled government AI service to make public administration accessible without typing or literacy barriers.
Lalitpur Metropolitan City Ward 3 — A 24-hour AI calling service for citizen complaints has gone live in Ward 3, Lalitpur Metropolitan City, built with Blue Oval. Callers ask about citizenship, land disputes, rubbish collection, health and senior citizen support, with issues logged for staff follow-up — part of a wider shift toward AI-driven public service delivery in Nepal.
Prime Minister's Office — Pakistan's federal government logs prime ministerial directives through the Prime Minister Office System (PMOS), unveiled by Shehbaz Sharif to demonstrate AI adoption in government. Digital workflows replace paper routines, with instructions monitored centrally. Officials call it the nation's first platform embedding public-sector AI use into governance.
Austrian Federal Chancellery — Austria's Federal Chancellery has begun rolling out GovGPT, a sovereign AI assistant for civil servants, starting in the Chancellery and moving department by department. Staff draft texts, condense long documents and build searchable knowledge libraries, with queries and uploads processed on Federal Computing Centre infrastructure instead of a public chatbot. The Ministry of Finance followed from 21 July and the consumer protection, care, health, social affairs and labour ministry on 28 July, targeting 180,000-plus federal employees.
Direction Générale des Impôts et des Domaines (DGID) — Senegal has scrapped paper tax stamps: DGID now channels purchases through SenTimbre, its digital tax stamps platform. A secure QR code on each stamp blocks the forgery that plagued the old paper-based electronic stamp duty system, part of the state's New Deal Technologique drive to digitise public services.
Djibouti National Meteorological Agency — Djibouti's National Meteorological Agency received MAZU-Djibouti 2.0, an upgraded AI weather alert system, on 17 July 2026. AI-driven forecasts merge satellite and surface data three days out at 3 km resolution (was 9 km), update six-hourly and offer one-click warnings. It backs December 2025's urban multi-hazard early-warning network.
Albanian Ministry of Interior (State Police) — Albania's Ministry of Interior has switched on AI policing in Albania spanning 873 sites: 12 Regional Police Directorates, 24 major cities, 28 border crossings and 20 Tirana schools all feed one Smart City command and control centre. Software there flags behaviour it judges dangerous or unlawful and raises an alert for officers to act on, and Interior Minister Besfort Lamallari insists the system amounts to more than cameras.
Senior Courts of Belize (High Court) — Belize's High Courts will use Lexis+ Protégé, an AI legal research and drafting tool, to speed up judgments. Judges and Belize Bar Association members trained on it at a 15–16 July High Court seminar hosted by Chief Justice Louise Esther Blenman and the Judicial Education Institute, following an April workshop on LexisNexis's AI tools.
Ministry of Energy and Business, Barbados — The Ministry of Energy and Business launched AI shopping assistant Ask Dealia on 15 July 2026 at the Warrens Office Complex, St Michael, Barbados. Infolytics built the app, the Department of Commerce and Consumer Affairs runs the service, and agent Dealia mines shoppers' conversations for bargains alongside live supermarket prices, promotions, events and personalised recommendations. Infolytics chief executive Quinn Weekes counts 400,000-plus products, 2,000 companies on the AI price-comparison tool.
Poder Judicial de la República Dominicana — The Dominican judiciary is piloting Microsoft Copilot in the Supreme Court of Justice's First Chamber and administrative areas, using it for routine tasks, drafting help and case-information queries. This AI legal document automation also covers standardising rulings, certifications and decisions on alimony, divorce and eviction cases, plus portal tools for text recognition, file anonymising and automatic hearing scheduling.
Eswatini Revenue Service (ERS) — Eswatini Revenue Service will trial AI tax audit risk detection across a 12-month span, singling out suspect filings for deeper checks. Brightwell Nkambule unveiled the push at the 4th Annual ERS Client Appreciation Day, calling AI-driven tax return risk flagging backup for auditors rather than their replacement.
Intendencia de Florida — Uruguay's Florida departmental government has launched Mi Portal Florida with FlorIA, an AI virtual assistant that guides users and answers common questions, on a single citizen portal for geolocated complaints, appointments and procedures. It aims to curb repetitive enquiries so staff can concentrate on resolving cases.
Regierungsrat, Canton of Zurich — The Canton of Zurich's government says it will make two piloted tools permanent across its administration: TranscriboZH, for AI transcription of audio and video, and KlartextZH, a plain-language text simplifier. Decided on 9 July 2026 (Regierungsratsbeschluss Nr. 749/2026), the step belongs to SynerKI, a cross-departmental project that will also build a central AI service platform for handling protected data. The Regierungsrat approved 2.76 million francs one-off and 3.87 million francs a year in running costs.
Ministerio de Tecnologías de la Información y las Comunicaciones (MinTIC) — Colombia's ICT ministry uses six AI, analytics and cloud systems from the ministry's IT office for AI in government work: generative AI manages citizens' petitions and complaints (PQRSD) and another tool runs contractor payments. The rest issue telecoms-operator charge sheets and debt-collection notices, step one of a broader public-sector AI programme.
National Board of Revenue, Bangladesh — Bangladesh's National Board of Revenue plans AI customs risk management, an Automated Risk Management System (ARMS) expected to nearly halve dwell time at Chattogram port. Replacing random picks, machine-learning screening will choose import containers for physical inspection by traders' compliance and risk profiles, under the World Bank-backed National Single Window project.
Government of Alberta — Alberta's Ministry of Technology and Innovation has built and deployed AI agents on Anthropic's Claude models to review more than 466 million lines of government code, after 18 months of tool-building. The review gave the province its first full picture of system health, and it plans to use the agents to replace 185 ageing systems with 16 applications it fully owns. Its 21 technical papers are open for other governments to reuse.
Cyprus Police (Traffic Police) — Cyprus has taken delivery of the first AI traffic cameras, which the traffic police say will go up on Griva Digeni Avenue in Nicosia in place of two units now sitting idle. The force says the equipment will flag unbelted drivers, helmetless riders, phone use at the wheel and other distractions, with a supplier still to finish preparatory work. These AI enforcement cameras will run as a trial first, judged on what the results show.
Organismo Judicial / Corte Suprema de Justicia — BDG Sociedad Anónima has been tasked with building AI cassation sentencing software for Guatemala's Supreme Court, following a Q89 mil 570 award from the Organismo Judicial that covers as many as 350 appeal files across a three-month span. Instead of deciding outcomes itself, the platform sifts through case records and turns out structured draft judgments, needing hookup with systems already running inside the judiciary. Magistrates alone hold the last word, fixed under Acuerdo 93-2026, leaving this AI-drafted court judgments tool as a support role only.
Luxembourg Ministry of Digitalisation — Luxembourg's Ministry of Digitalisation put an AI chatbot government services tool live on Guichet.lu, the country's citizen portal, on 6 July 2026, under the AI4Lux campaign led by Minister Stéphanie Obertin. Residents can now type queries in whichever language they speak and get routed automatically among 2,000 bureaucratic procedures, rather than hunting for the right form by name. The assistant, an early example of conversational AI for public administration, runs entirely on government-owned servers.
Câmara Municipal de Portimão — Portugal's Portimão city council says it will adopt a digital planning platform, and its online service will use AI to check documents the moment residents submit them, as all building licensing moves online. Municipal staff training has begun, and the platform is arriving in phases, expected to go live within months.
Instituto Nacional de Ciencias Forenses de Guatemala (Inacif) — Guatemala's National Institute of Forensic Sciences (Inacif) has for the first time identified a deceased person using IdentIA, AI tattoo matching software that analyses and classifies tattoos for body identification. A 24-year-old woman had lain unidentified in the Guatemala City morgue for about a month until the system matched her tattoos to her mother's ante-mortem interview; fingerprint comparison then confirmed her identity.
Gobierno de la Ciudad de Buenos Aires — Buenos Aires city government launched BAX, an AI city services app putting cultural activities, information requests, complaints, appointments and procedures in one place. BAX's conversational AI takes text, voice or images, replacing forms and menus scattered across apps and sites.
Ministry of Industry and Trade — Yemen's Ministry of Industry and Trade has cleared a smart programme running on artificial intelligence and reaching the public through WhatsApp, a decision taken at a meeting in Aden chaired by Minister Mohammed Al-Ashwal. Citizens will reach the department's offerings — AI government services delivered by messaging app — and the smart public-service programme will feed the RASAD platform, where complaints about trading breaches are lodged and chased, shortening handling times.
Skatteverket and Försäkringskassan — Sweden's tax agency Skatteverket and social insurance agency Försäkringskassan have begun trial operation of AI-verkstaden, a shared AI workshop that helps agencies, municipalities and regions build and use AI. Ten pilot organisations joined first, including Region Halland and Mörbylånga municipality, testing redaction, transcription and digital-assistant AI services. The pilots were picked to span different sizes and levels of technical maturity, the agencies said.
South African Revenue Service — The South African Revenue Service selects 100% of verification cases and 88.41% of complex audit cases by automated AI tax risk assessment, as machine learning checks declarations against third-party data from banks, employers, retirement funds, medical schemes and crypto reporting frameworks before returns are processed. More than 1.9 million taxpayers were auto-assessed, and about R8 billion refunded within 72 hours. Communications head Siphithi Sibeko says human oversight and governance processes review the models' risk indicators before any enforcement.
State Customs Committee of Uzbekistan — An AI customs robot called AiRo is now working in Tashkent airport's international terminal, running software developed with Uzbekistan's State Customs Committee. Passengers can ask about bringing phones, medicines, jewellery or currency in or out, prohibited goods and how the red and green channels work. The committee wants the interactive robot to inform passengers fast, ease officers' work and prevent violations.
State Election Commission — Croatia's State Election Commission has opened GlasAI, an AI chatbot for elections fielding public questions on ballots, referendums and the commission's remit, built with the Ministry of Justice, Public Administration and Digital Transformation. Reachable at izbori.hr under a stylised ballot-box-and-headset icon, the electoral assistant has replied 24 hours a day, 7 days a week since 1 July 2026.
Agency for Social Insurance and Pensions of Tajikistan — Tajikistan's Agency for Social Insurance and Pensions has made facial recognition pensioners checks mandatory for all recipients, after a nationwide re-registration logged nearly 877,000 beneficiaries into a refreshed electronic database. The drive exposed fraud, including payments still drawn on behalf of deceased recipients, and forms part of a wider shift toward biometric identification pension systems. The agency is now building remote online verification so citizens abroad can complete facial scans without returning home.
Ministry of Municipalities Affairs and Agriculture (Bahrain) — Bahrain's Ministry of Municipalities Affairs and Agriculture is trialling AI surveillance cameras at some public sites to catch littering, vandalism and other offences, it said on 30 June 2026. Control rooms analyse the smart cameras' feeds; rollout will widen yearly across the ministry's 200-plus parks, gardens, beaches, seafronts and walking tracks. Guards cannot be stationed everywhere, the under-secretary noted.
National Centre of Meteorology (UAE) — The UAE's National Centre of Meteorology introduced agentic AI in weather forecasting, with two AI assistants in its operations rooms as a roadmap's first step. 'Al Rased' raises alerts from continuous monitoring of weather stations, radars, satellites and seismic and air-quality networks; 'Forecaster's Assistant' weighs global model outputs against numerical-model data and writes draft bulletins. Humans keep the final say, per the director general.
Ministry of Justice of Türkiye — Türkiye's Ministry of Justice has given judges and prosecutors the CBS Organisation Prediction Project, an AI case analysis tool that reads case-file data to predict which organisation a file may link to. The ministry's Directorate General of Information Technologies calls it a digital assistant meant to shorten trials and support decision-making.
Ministry of Youth Empowerment, Employment, Sport and Civil Service (Mauritania) — Mauritania's Ministry of Youth Empowerment, Employment, Sport and Civil Service introduced Morshed, billed as the country's first AI government assistant. The bilingual assistant answers in French or Arabic on the ministry's job offers, civil-service rules and youth and sport schemes, drawing on its databases and official documents and sourcing every reply.
MASGZ Brandenburg — Brandenburg activated an AI in immigration services tool within 'Aufenthalt Digital' from 26 June 2026, expanding a six-month Dahme-Spreewald pilot statewide. The system flags missing or contradictory paperwork before submission, spanning permits for employment, training, family reunification and settlement. Some 350 German offices already run the portal, now enhanced with automated document verification for visa applications.
Macao Public Security Police Force — Face-scan border clearance in Macao now spans 204 automated channels at every joint checkpoint, after officials activated gates at Qingmao Port and the Zhuhai-Macao hall of the Hong Kong-Zhuhai-Macao Bridge Port on 26 June 2026. A Hengqin Port trial in November 2025 preceded this wider, paperless border-crossing launch, open to registered Macao residents, Hong Kong permanent residents and mainland travellers aged 14 or above.
National Accountability Bureau — Pakistan's National Accountability Bureau has set up an AI investigation system to support white-collar crime investigations. National Centre of Artificial Intelligence chairman Dr Yasar Ayaz says the platform is built to assist investigators, and it should make investigations more effective once running; the bureau has not reported it in live use.
Crisis Management Center of North Macedonia — North Macedonia's Crisis Management Center (CMC) has agreed with NATO to implement NICS-2, a multi-year AI crisis response project delivering an incident command system, AI-driven, for cross-border emergencies. Under NATO's Science for Peace and Security Programme, this AI-assisted emergency response effort targets stronger national crisis management capabilities via AI-enhanced coordination and real-time decision-making.
Poder Judicial de Río Negro — The Poder Judicial de Río Negro has brought AI anonymisation of court rulings into its Puma case-management system, ending the manual marking of personal data by staff. Developed by the judiciary's own Área de Informatización de la Gestión Judicial, it scans each document for names, identity documents, CUIL numbers and case-caption details needing protection and proposes an anonymised text every court office can check and adjust. It is optional for most documents but central for rulings due for publication.
Judicial Service of Ghana — Ghana's Judicial Service will equip all 445 judges and magistrates with Eskwai Clerk, an AI judicial assistant built alongside Kwame AI for legal research, reviewing documents and writing rulings and judgments. Training will frame the AI helper for the bench as an aid that leaves judicial discretion intact.
Metropolitan Police — Live facial recognition policing expands to London's West End and Soho by Christmas 2026, the Metropolitan Police says, fixing cameras to street furniture ahead of six further rollouts in 2027. A prior Croydon trial scanned 470,000 faces over six months, yielding 173 arrests and one false match, evidence now cited to justify wider facial recognition surveillance.
Bundesagentur für Arbeit (Federal Employment Agency) — Germany's Federal Employment Agency (BA) is rolling out AI customer service for public agencies nationwide, extending its Antwortvorschläge reply-suggestion tool to every SGB III Service Centre, with completion due by 30 September 2026. The system drafts responses to mailbox queries, but staff must check and approve each one before sending. This nationwide push follows a longer internal test phase, with automated drafting still requiring human sign-off throughout.
Department of Information and Communications Technology — The Department of Information and Communications Technology says generative AI for public servants starts with over 50,000 staff getting Google Cloud's Gemini Enterprise app under a multi-year partnership, then more than 200,000 government workers within 18 months. It will also build AI agents speaking local languages to help citizens with disaster aid guidelines, health centre appointments and administrative procedures.
Ethiopian Artificial Intelligence Institute — One app now handles Ethiopia's AI public services: the Unified MESOB Platform, built by the Ethiopian Artificial Intelligence Institute, merges federal agency functions into a single web and mobile portal. Abiy Ahmed opened it, while Worku Gachena credits it with cutting red tape and sharpening openness, part of a broader shift toward digitised citizen access nationwide.
Government of Mauritius — Mauritius's National Electronic Licensing System (NELS) will gain an AI chatbot for licensing, guiding businesses and citizens through licensing steps in several languages. The 2026/27 budget also funds AI tools for the Financial Services Commission and a new Healthcare Innovation and Artificial Intelligence Unit to judge clinical AI rollout.
City of Toronto — The City of Toronto says it will launch its own AI chatbot for general public enquiries on its 311 site, leaving phone agents for harder issues. Part of a redesigned 311 website due by the end of 2026, it could explain how to file a request or how long a pothole repair takes. The city says 311 gets 1.5 million calls a year, about 60 per cent general enquiries.
Ministry of Interior — 487 AI-equipped CCTV cameras now watch Phnom Penh, an AI policing push the Ministry of Interior calls its Lancang-Mekong Regional Smart Policing Development project, unveiled at a ceremony Deputy Prime Minister Sar Sokha turned up for. Footage streams into one command hub for quicker emergencies and non-stop watchfulness, powered by facial recognition cameras and instant number-plate capture. Secretary-General You Xiaowen of the Lancang-Mekong Integrated Law Enforcement and Security Cooperation Centre built it, vowing coaching for Cambodian officers and a wider grid.
Dorset Council — Faculty, an AI firm working with the UK government, built the agents now being piloted by Dorset Council to assist planning officers on minor applications, reviewing documents and issuing recommendations. This AI planning agents trial aims to halve average processing times, cutting the wait from eight weeks to four. Two other councils are running the same test, with AI in planning application processing meant to ease backlogs.
French Southern and Antarctic Lands (TAAF) — Sorting terrain across Amsterdam and Saint Paul Islands into 55 distinct categories at 92.6% accuracy, France's TAAF (French Southern and Antarctic Lands) now runs AI habitat mapping through its Cartovege2 machine learning tool, drawing on Pleiades pictures shot from orbit at 50cm resolution. Built alongside the University of Rennes and the French Polar Institute, the system converts raw imagery into usable conservation data, advancing satellite habitat classification from research project to working instrument. Staff had already leaned on it to gauge harm after fire tore through half of Amsterdam Island in January 2025.
Open Knowledge Foundation Nepal — Open Knowledge Foundation Nepal added a GPT-4.1-mini conversational assistant to Open Data Nepal, an AI open data portal holding more than 630 datasets from 48 publishers in 24 categories. Visitors ask in plain English to find datasets, decode column meanings, produce charts. The AI-assisted portal covers governance, health, education and environment.
French government (Notre IA plan) — The French government will bring generative AI for public advisers to its 8,000 France services staff by autumn 2026 with a tool built on Aria, a chatbot from the Directorate of Legal and Administrative Information (Dila). It is part of the Notre IA plan unveiled at Bercy, which also foresees a general conversational tool for some one million state employees, phase one costing an extra €750,000 in Mistral licences and computing power.
Gobierno de la Provincia del Neuquén (Dirección de Fauna) — Neuquén's governor Rolando Figueroa says the province's wildlife wardens will get 16 AI anti-poaching drones with thermal cameras from 1 July to fight illegal fishing and hunting. They also receive 60 tablets for digital enforcement records and Starlink antennas for remote areas, working jointly with the provincial police.
Federal Ministry for Digital and State Modernisation (BMDS) — Germany's Federal Ministry for Digital and State Modernisation has open-sourced SPARK Workflow, an AI planning approvals system any government level can now adopt. Agents flag gaps or contradictions in submissions and draft rulings pending human approval. Trials run at the Federal Highway Authority, Hamburg's hydrogen core grid permitting, and North Rhine-Westphalia's emissions clearances, pointing toward wider AI-assisted permitting.
Ministry of Innovation and National Development — Sebas Bastian, minister of innovation and national development, told the 2026-2027 Budget debate that SmartGov, a package of AI tools for civil servants, will absorb briefs, research, correspondence and paperwork. The Government will not lay software over today's forms, but rewire the workflow beneath, so a task that used to pass across a desk becomes automatic. A secure national identity card anchors the joined-up state these public sector AI tools serve.
Uganda government — Ugandans now get AI government services Uganda-wide through UGGov Agent, one channel to reach state bodies, follow public project progress and get help in 16 local languages over WhatsApp, Telegram and USSD. AfroSoft IT Solutions built this citizen-facing government AI assistant with presidential backing, drawing data from various ministries, departments and agencies for owners of smartphones and basic feature phones alike, sparing them an office trip.
Ministry of Justice, Public Administration and Digital Transformation — Croatia's Ministry of Justice, Public Administration and Digital Transformation has switched on an AI virtual assistant for public services at the heart of the state portal gov.hr, delivered inside the National Recovery and Resilience Plan scheme building one contact hub for every e-public service. Residents phrase a query the way they would put it to a clerk across a counter — renewing a passport, enrolling a child, registering a move — and the government service chatbot hands back an ordered reply, pointing at the matching e-service. Availability is 24 hours a day, 7 days a week, with no personal data demanded or retained.
Tax Inspection Board (VDK), Ministry of Treasury and Finance (Türkiye) — Türkiye's Ministry of Treasury and Finance says its Tax Inspection Board will launch the AI-supported Bilinç Platform this year, letting taxpayers see and fix board-identified tax risks before any audit or penalty. Phase one gives AI tax risk scoring, combining over 150 rule-based tax analyses with machine-learning scenarios, and inspectors may be assigned to taxpayers seeking help.
Municipal Affairs Bureau (IAM) of Macao — Macao's Municipal Affairs Bureau (IAM) applies AI food safety risk assessment via a new Food Safety Risk Monitoring and Assessment Network Prediction and Early Alert Platform. Built since 2023 with Chinese mainland research institutes, it tracks 109,907 pre-packaged food products through AI-assisted food safety monitoring. Since 2026, six alerts have gone out: 148 SMS texts, 374 emails.
National Employment Agency (ANAPE) — AI job matching now runs inside Mauritania's National Employment Agency (ANAPE), pairing jobseekers' abilities against open vacancies far faster than before. Minister of Youth Empowerment, Employment, Sports and Civil Service Mohamed Abdallah Ould Loly cast this as one piece of a larger campaign to modernise public-sector hiring channels through digital means, alongside branch offices already stretched to eight localities nationwide. The shift signals broader momentum toward AI in public employment services across the nation.
Ministry of Justice — Taiwan's Ministry of Justice launched an AI in legal administration project, the Ministry of Justice Exclusive Large Language Model, described as the island's first sovereign AI system built specifically for legal work. Minister Cheng Ming-chien said the legal domain AI model targets the administrative burden facing prosecutors, who together process more than 2 million criminal cases each year, freeing them to concentrate on analysing cases and flagging risk. The initiative applies government AI to prosecutorial functions rather than private law practice.
Municipalidad de Santa Ana — Costa Rica's Municipality of Santa Ana has started a pilot in which two AI agents for public procurement will help departments draft market studies and technical specifications for tender documents. A workload study showed departments sinking time into SICOP e-procurement paperwork; municipal police, parks, culture, public works and purchasing are all taking part.
Agência Nacional de Saúde Suplementar (ANS) — Brazil's supplementary health regulator, the ANS, is developing Iara, an AI claims review tool to assess insurers' challenges to SUS treatment charges for their members, with analysts checking every recommendation. A first module is due by end-2026 and full rollout in 2028; developer Paipe won a public tender.
Ministry of Interior of Serbia — Serbia's Ministry of Interior runs NEC's NeoFace Watch and NtechLab's FindFace, ministry sources told BIRN, extending facial recognition surveillance already built on thousands of Huawei cameras with biometric capability. Russia and Iran use the same software. Rights groups warn this biometric identification system lacks legal grounding, since biometric provisions were left out of the draft Law on Police.
State Customs Service of Turkmenistan — AI in customs and trade compliance is now live on Turkmenistan's Single Window for Export-Import Operations portal, where the State Customs Service added an assistant handling several administrative tasks at once. It works out HS commodity codes, warns which permits are needed, and gives instant duty-rate estimates, all in Turkmen, Russian or English. Built with UNCTAD backing, this AI-powered trade portal aims to speed up cross-border processing.
King Abdulaziz City for Science and Technology (KACST) and General Directorate of Public Security — Saudi Arabia's KACST, working with the General Directorate of Public Security, unveiled Roya, delivering AI crowd monitoring Hajj coverage at the holy sites. Merging GIS with orbital pictures, the tool measures congestion, follows pedestrian and vehicle movement, and pinpoints heat zones across some 2,000 images plus 126 aerial missions spanning Hajj and Umrah, with automated crowd density tracking spanning raw capture through live output.
Ministry of Management and Innovation (MGI) — Brazil's GOV.BR portal answers up to 75% of education and health questions through an AI chatbot for citizen services, across 157 thousand chats. Minister Esther Dweck's INSPIRE project built the government chatbot; ministries record 182 AI use cases.
National Disaster Management Organisation (NADMO) — Ghana's National Disaster Management Organisation introduced a WhatsApp tool for AI disaster alerts on 27 May 2026, developed by UNESCO with Government of Japan backing. The flood and drought chatbot sends forecasts, warnings and advice, and residents flag hazards and updates to disaster agencies. NADMO's Director-General wants responders using it daily.
Liechtensteinische Landesverwaltung (LLV) — LLV, Liechtenstein's State Administration, broadened its AI chatbot government website tool LiBot to cover all of www.llv.li, moving beyond its earlier commercial-register-only scope. The assistant, an AI-driven virtual assistant for public sector services, resolves common queries at any hour and steers visitors toward the right information, built on an AI model tuned for clear, accurate replies. Officials frame it as part of a wider digitalisation strategy.
Ministry of Finance (UAE) — The UAE Ministry of Finance has introduced AI call centre analytics that record and assess calls, reading caller sentiment and showing unhappy calls red and resolved ones green on a managers' dashboard. It also logs call length, resolution time and each handler's performance score, and nudges agents mid-call to swap casual phrases for formal, service-minded language.
Government of Guyana — Guyana's President Irfaan Ali says a mobile AI traffic enforcement unit is now patrolling roads, automatically spotting vehicles with unpaid tickets so they can be stopped. The surveillance unit's pilot days earlier caught 250 vehicles with outstanding tickets.
Federal Ministry of Communications, Innovation and Digital Economy — Nigeria's Federal Ministry of Communications, Innovation and Digital Economy launched GovGuide Nigeria, an AI chatbot for government services answering citizens by voice or text in English, Hausa, Igbo and Yoruba. Local firm Publica AI built the voice-and-text assistant on Meta's open-source Llama models with the National Centre for Artificial Intelligence and Robotics.
Ministerio de Desarrollo Social (MIDES) — MIDES runs an AI parenting support chatbot on WhatsApp, Asistente Virtual de Crianza, answering Uruguayan parents on pregnancy and upbringing questions. This WhatsApp chatbot for childcare guidance draws on Uruguay Crece Contigo, the Ministry of Public Health, PAHO and UNICEF, and flags vulnerability signs to telecare staff.
Town of Otrokovice — Otrokovice, a Czech town, has switched on an AI assistant for council websites so residents can find things faster. Tucked behind a corner icon, the tool takes questions without sending anyone away from the page, combing municipal pages and other material for dependable, up-to-date replies on services, official business and what is happening locally. A spin-off of the Faculty of Mechatronics at the Technical University of Liberec supplies the municipal information chatbot, which the Constitutional Court also runs.
General Directorate of Customs (Tunisia) — Tunisia's General Directorate of Customs added an AI module for AI customs risk targeting to the national customs selectivity system that picks declarations for control, it said on 16 May. Its machine-learning algorithms sift customs big data, automatically flagging high-fraud-risk operations to sharpen checks and speed compliant traders' goods.
Government Digital Service — The Government Digital Service has released GOV.UK Chat inside the GOV.UK app, letting residents pose everyday questions and get replies drawn from verified public-sector sources. This AI government chatbot ran as a soft launch since March, processing over 15,000 queries from more than 7,800 distinct users. It now moves to full public availability, an AI tool for public information queries built into the app.
High Court in Prague — Prague's High Court has handed citizens Čenda, an AI legal guidance chatbot running the Claude model on the Lovable platform. Čenda quotes no statute, sketching what a person divorcing or mired in debt might try, sourced from justice.cz and gov.cz. Issued alongside the automated legal help tool were 18,121 civil and 6,314 insolvency judgments stretching to 2020, the pair billed 800,000 crowns.
Cyprus Fire Service — Cyprus Fire Service runs four AI wildfire detection units reaching roughly 80 per cent of unpopulated ground in the government-controlled zone. Brigade chief Nikos Logginos ranks the units central to the 2026 summer plan, saying each spots a flame or plume no bigger than a square metre. Another 13 are earmarked for industrial estates on an €800,000 budget, while taking AI smoke detection into open country and woodland awaits finance ministry approval.
Supreme Court of India — The Supreme Court of India has launched Su-Sahayak, an AI court chatbot, alongside a 'One Case, One Data' integrated case management platform. Built by the National Informatics Centre with the Supreme Court Registry, the digital assistant offers court guidance, procedural help and access to judgments, orders, cause lists and case status.
Gobierno de la Provincia del Neuquén — Neuquén's provincial government runs an AI consumer protection chatbot answering consumers day and night. A panel of AI agents reads each query, checks procedural requirements and legal guidance, and replies with step-by-step instructions.
Programa Juntos (Ministerio de Desarrollo e Inclusión Social) — Peru's Juntos social programme, of the Ministry of Development and Social Inclusion, launched MarIA, an AI chatbot for social programmes answering beneficiaries on WhatsApp day and night in Quechua and Spanish. Built in-house, the WhatsApp assistant passes cases needing personal attention to a user-support technician.
Israel Courts Administration — Israel's courts administration is extending 'Chat HaMishpat', an AI assistant for judges, from pilot to every judge's and legal assistant's chambers, 1,650 people. The court AI tool, built on Google's Gemini engine by the administration's digital technologies division, handles only case documents securely, summarising pleadings and testimony, analysing documents and making tables. It is meant to lift technical work from judges without touching judicial discretion.
Iraqi Ministry of Planning — Iraq's Ministry of Planning has used AI census analysis to turn the huge dataset from the 2024 population census into indicators for development plans and investment priorities. Spokesman Abdul Zahra al-Hindawi adds that automated contractor naming and classification now runs within Central Organisation for Standardisation and Quality Control services.
Tribunal Superior de Justicia de Querétaro — The Tribunal Superior de Justicia de Querétaro used Sonia, its self-built and self-hosted generative AI, to help write two final rulings, in a commercial enforcement case and an uncontested divorce, reportedly a first for generative AI in Mexican courts. Court president Braulio Guerra Urbiola says around 17 million pesos went on a dedicated physical server to keep litigants' data in house, with six months of training on family and commercial law. Specialists supervise every Sonia document and the presiding judge assesses it before issue, keeping legal responsibility with the judge.
Ministry of Public Administration of Montenegro — Montenegro's Ministry of Public Administration has rolled out an AI chatbot for public services, live on the Government and eUprava portals, sparing citizens paperwork to find information. Minister Maraš Dukaj called round-the-clock, 24/7 access the key benefit of this AI-powered government assistant. Still in early pilot stage, it will gain new capabilities as feedback arrives.
Gobernación de Cundinamarca — Residents of Cundinamarca can now lodge PQRSDF requests over WhatsApp, using an AI virtual assistant government services channel called Policarpa. Linked to the Cóndor document system, it issues a filing number holding the same legal weight as in-person paperwork, an AI-powered public service assistant cutting the need for office visits.
Comunidad de Madrid — SOL, a new AI public services chatbot from Madrid's Consejería de Digitalización, now serves 7 million residents around the clock. Built on Microsoft Foundry with Azure OpenAI Service, it answers voice and text queries in Spanish, English and Romanian, using over 70,000 official URLs from the regional government and transport authority. Atos, Bravent and T-Systems helped build the digital assistant.
Lagos State Government — The Lagos State Government's Lagos Emergency Responders Management System went live on 27 April 2026, connecting the state ambulance service, LASTMA and the fire service to the Lagos State Command and Control Centre. AI emergency call screening checks reports live, acting as a prank-call filter against false alarms, and pinpoints callers automatically.
Government of The Gambia — The Government of The Gambia has commissioned the National Misinformation and Disinformation Response Centre, an AI-powered misinformation detection platform under the Ministry of Information that scans online conversations, spots emerging false-content trends and verifies claims against trusted sources, functioning as an AI-driven disinformation monitoring hub. Officials call it West Africa's first such centre, built by Gambian firm Dundal Systems with ECOWAS backing and unveiled at the Sir Dawda Kairaba Jawara International Conference Centre.
Parliament of the Cook Islands — The Parliament of the Cook Islands switched on two AI agents in parliament on 24 April 2026: INA, a public-facing chatbot on the parliament website, and Tumutoa, an inward-looking staff guide to procedure plus institutional knowledge. Clerk of parliament Tangata Vainerere billed the institution “officially AI-enabled” yet insisted work remains “human-led, not AI-led”. Pacific Group AI built this parliamentary AI, which founder Tayla Jayne Beddoes calls the Cook Island government’s first. Roughly five AI tools reached staff weeks earlier; the United Nations Development Programme paid via Pacific Parliamentary Effectiveness Initiative Phase 3, a five-year Government of New Zealand scheme.
Alcaldía Mayor de Bogotá (Consejería Distrital TIC) — Bogotá's District ICT Office has switched on an AI child protection system that follows more than 640,000 under-sixes individually to foresee vulnerability, violence included. It merges education, health and social-integration data, combines a predictive model with case management, and, after a Kennedy pilot with territorial teams, is spreading citywide in stages.
Fiscalía General de la República de El Salvador — El Salvador's Attorney General, Rodolfo Delgado, says his office is building a sovereign AI platform with WideLabs and NVIDIA Brasil for 24-hour AI criminal complaint intake. People will file by voice or text on WhatsApp or the web, and each complaint will be structured and sorted automatically, arriving analysis-ready with the prosecutor in charge.
Government of Ukraine — Ukraine's Ministry of Digital Transformation is trialling AI policy impact analysis, software that predicts how draft acts will shape outcomes for citizens, businesses and other groups, labelling each positive, neutral or negative. Acting minister Oleksandr Bornyakov described it as an early-stage experiment, aimed at handing officials quicker, evidence-grounded judgement than manual review offers. This AI-driven policy assessment lacks a set launch date beyond the current pilot.
Municipality of Tripoli Centre — Libya's Municipality of Tripoli Centre is building BaladiyaBOT, an AI civic assistant for municipal services, backed by GIZ and the EU. It will pilot locally before wider rollout as a local government chatbot.
Murang'a County Government — Murang'a County's MyMurang'a app puts every county department in one place, adding an AI legal chatbot for on-the-spot answers and pro bono referrals. Residents can book clinics, search land records and seek business permits or bursaries; farmers get AI tips on pests and plant diseases. Governor Irungu Kang'ata unveiled the legal-AI app in Kenol.
Superintendencia de Electricidad y Combustibles (SEC) — Chile's electricity and fuels regulator, the SEC, uses AI import inspection to catch uncertified electrical products entering the country sooner. Reading import data, raising alerts and ranking controls by risk, the tools lifted the certified share from 88% to 95% and cut months-long detection times, the regulator says. An SEC AI assistant also answered over 60% of queries during electricity-subsidy applications.
Georgia Revenue Service — Georgia's Revenue Service uses AI VAT refund automation to return over GEL 1 billion to the economy yearly. A risk-scoring system checks payment history, supplier invoices and sector benchmarks against each monthly filing; if nothing flags, funds reach the firm's account within five to seven working days. This tax refund automation skips manual review for low-risk cases, though flagged filings face further scrutiny.
Ministry of Local Development and Environment (Egypt) — Egypt's Ministry of Local Development and Environment will introduce AI drone monitoring with geospatial analysis under a Drone Tech agreement. The aerial surveillance will track sources of pollution such as the seasonal 'black cloud' from crop burning, plus illegal building, waste dumping and encroachment on nature reserves, aiding environmental inspection. Governorates will be covered in phases.
Commonwealth of Pennsylvania — Over 3,000 Commonwealth employees across 35 agencies in Pennsylvania, United States, now have access to ChatGPT Enterprise under an expansion by the Shapiro Administration, part of a wider state government AI deployment, with 6,500 more staff required to complete training on safe and responsible AI use before gaining entry. This follows a 12-month pilot in which 175 workers spanning 14 agencies reported cutting an average of 95 minutes from their daily workload while drafting messages, condensing documents, researching topics and generating ideas, an early sign of AI in state government paying off. Officials note human oversight stays central to the effort, and flag that the 95-minute saving is a self-reported figure, not independently verified.
International Water Management Institute and Limpopo Watercourse Commission — Botswana, Mozambique, South Africa and Zimbabwe gained two fresh AI systems for overseeing the Limpopo River Basin on 15 April 2026, once the International Water Management Institute and the Limpopo Watercourse Commission switched on WaterCopilot and the Limpopo Digital Twin. Built alongside Microsoft, WaterCopilot responds to officials' everyday-language queries by drawing directly from hydrological records, serving as a conversational AI water management tool. Meanwhile its counterpart keeps continuous watch over the basin, feeding climate adaptation efforts and joint cross-border choices on shared water resource planning.
Tokyo Metropolitan Government — The Tokyo Metropolitan Government has fully launched A1, a generative AI platform for government staff, to about 60,000 metropolitan officers after a trial from September 2025. Built by GovTech Tokyo with the Digital Services Bureau, it lets officers build and share no-code AI apps, from contract-specification drafting to AI adoption checkpoints and assembly answers drawn from transcripts.
Financial Administration of the Czech Republic — The Ministry of Finance and the Financial Administration have opened a beta of an AI tax return filing tool, which the ministry calls a first for Czech public administration. Upload a payslip and the software picks out the figures required, dropping each into the right box, so nothing needs retyping and no row numbers need finding. Automated tax form completion covers only wage earners for now, including people paid by several employers and students earning on the side.
Municipality of Ratiboř — Ratiboř, a village in the Zlín Region, now fields enquiries through an AI mayor — a digital twin of Martin Žabčík, built over two years by Digiregion with AI Produkce. The likeness appears on a video screen in the council building's entrance hall, hooked into the village's own knowledge store, so callers learn of roadworks or planned spending. Because the digital council avatar talks day and night, residents wary of laptops and mobiles gain too.
Fiji Revenue and Customs Service (FRCS) — Fiji Revenue and Customs Service now runs an AI chatbot for tax enquiries, named Masi, on its new taxpayer website. Corporate Services Director Shavindra Nath said the conversational AI system handles both basic and complex questions 24 hours a day, seven days a week. Automating routine responses frees staff to focus on harder cases, and the tool doubles as an automated taxpayer support tool improving overall service.
Direction régionale des finances publiques de la Guyane — Direction générale des finances publiques in French Guiana now runs an internal AI writing tool, "langage clair" (plain language), simplifying taxpayer communications. Officers submit drafts, which the system rephrases into clearer wording before release, backing this AI-assisted tax communication push. It supports staff through the 2026 tax campaign, lifting plain-language government correspondence.
Government of Zimbabwe — Zimbabwe will launch AI fraud detection financial services through the Zimbabwe AI Cyber Shield, a national programme under the country's National AI Strategy. ICT Minister Tatenda Mavetera outlined a 12-month rollout at the Cyber Fraud and AI Summit, alongside training for 10,000 cybersecurity professionals in AI-powered fraud prevention and a legal framework for ethical AI use in cyber defence. Mobile money fraud driven by AI drains US$30 million from Zimbabwe each year, the reason behind the push.
Burkina Faso Ministry of Security — Burkina Faso's Ministry of Security started a nationwide video-ticketing trial in Ouagadougou on 7 April 2026, with high-definition AI traffic cameras spotting red-light, seatbelt, phone and speeding offences. Each flagged vehicle's owner gets an SMS with a link to pay; camera-based enforcement was scheduled to run fully from 1 August 2026.
Zhogorku Kenesh (Parliament of Kyrgyzstan) — Kyrgyzstan's Parliament has piloted E-Kenesh, a tablet-based platform replacing paper documents for MPs. Speaker Marlen Mamataliev introduced the tool, whose AI bill summarisation condenses lengthy draft laws and helps legislators frame questions under debate. Materials appear three days early, with staff filing documents through AI-generated legislative summaries online.
Ministry of Technology and Communications — LaoAI, a new government AI assistant for civil servants, launched alongside G-Chat v2 in Laos on 6 April 2026. Built by the Ministry of Technology and Communications, the tool drafts and summarises documents, retrieves regulatory information and translates text across all ministries, supporting AI for public administration nationwide. Deputy Prime Minister Saleumxay Kommasith attended, and the minister pressed every state body to adopt it for daily work.
Controladoria-Geral da União (CGU) — Fala.BR, Brazil's federal ombudsman platform run by CGU, now offers AI citizen service, automatically sorting complaints by reading each submission to determine its manifestation type. A planned upgrade will let the tool recommend the receiving body, helping automate complaint classification.
Občina Sveta Trojica v Slovenskih goricah (Municipality of Sveta Trojica) — Klara, an AI digital secretary, has fielded calls for Sveta Trojica municipality since 2026. Unveiled by Mayor David Klobasa on Easter Sunday and live from that Friday, the tool grasps what callers want, replies, and auto-drafts written summaries as official records. Council goals cited included speed, order and openness, though human contact, officials insisted, stays central: artificial intelligence for municipal call handling will not replace it.
Agencia Estatal de Administración Tributaria (AEAT) — Spain's AEAT tax authority deploys AI forensic analysis tax agency tools, rolling out 500 licences of Cellebrite Pathfinder to sift unstructured digital evidence—emails, images, videos—in open cases. The tool flags patterns and anomalies in seized data without replacing human judgement, an AI-powered digital evidence review layered atop inspectors' work. AEAT paid roughly €326,000 for the software in 2023; fiscal law experts warn its reach may extend beyond verifying income to mapping a taxpayer's social and family circle.
Ethio Telecom — Ethio Telecom's new teleSign platform verifies users through AI identity verification, matching facial biometric verification via liveness-detection video against the Fayda National Digital ID instead of an office visit. Citizens and diaspora Ethiopians use it to confirm paperwork, assign or withdraw legal representation, and apply for professional licences from a phone or browser. Ethiopia's Ministry of Justice, Ministry of Foreign Affairs and the Federal Documents Authentication and Registration Service partnered on development.
Conselho Administrativo de Recursos Fiscais (Carf) — Brazil's federal tax appeals council, Carf, published ordinances in the Official Gazette on 30 March creating Iara, an AI case-law search tool covering its rulings since 2012 to help councillors draft votes, plus generative AI rules. For 30 days access was restricted to councillors and curation staff in a pilot group; users must review outputs and officials remain accountable.
Abu Dhabi City Municipality — Abu Dhabi City Municipality now has 27 AI inspection vehicles under its Unified Inspection System. Using cameras, detection systems and automated analysis, they spot visual problems in public spaces unaided and send locations and data to inspection teams, reducing manual inspections. Recorded visual-distortion cases climbed from 3,632 in 2024 to 18,851 by end-2025, which the municipality puts down to wider coverage rather than more problems.
Government of Mauritania (National Competitions Committee) — Mauritania held its first entirely digital public recruitment exam, with AI recruitment screening verifying online applicants' data, says Prime Minister El Moctar Ould Djay. Of 1,459 applications, 513 were accepted; facial recognition checked identity, the test ran on camera-fitted tablets and marking was automatic. The National Competitions Committee planned further digital exams for IT engineers and specialist doctors.
Greek Ministry of Digital Governance and AI — Greece's Ministry of Digital Governance switched on pilot AI traffic cameras at eight high-risk points in Attica on 28 March 2026, certifying violations through the traffic police system and notifying drivers via gov.gr. The first 130 digital fines, mostly for red-light running and helmetless motorcyclists, went out within days; three ministries share the programme: digital governance, citizen protection and transport.
Judiciary of Uganda — Uganda's Judiciary now runs AI legal research software to sift case files, cutting the time judges once spent digging through records. Uganda Legal Information Institute (ULII) added a matching tool that reads filings and condenses long rulings into short summaries, an early example of AI-powered case analysis in the courts. Paper filing disappears by 1 July 2026 under the Electronic Court Case Management Information System (ECCMIS).
Gibraltar Parliament — Gibraltar Parliament has opened the whole legislative archive — every Hansard volume back to 1969 — through a new AI Hansard search built for public use. Beyond debate transcripts the service reaches Speakers' Rulings and Standing Orders, turning precedent hunting that once required manual research into a single query. This AI-assisted parliamentary research assistant condenses debates neutrally, yet cannot draft communications, speeches or anything original. The interface is already running on parliament.gi.
Independent Authority Against Corruption (IAAC) of Mongolia — Mongolia's IAAC (Independent Authority Against Corruption) has turned to software for AI government monitoring, piping every state order and regulation through the tuss.io platform to hunt for legal breaches, wasted procedure and hidden conflicts of interest. Automated review of public-sector decision-making has already touched 388 orders, forcing 152 into revision or the scrap heap.
Canton Ticino — Ticino's public administration has rolled out an AI translation for government staff tool built on Switzerland's open-source Apertus language model, with roughly 100 civil servants trialling it in the initial rollout phase. Artificialy, the Ticino-based firm behind the system, designed it to convert text across German, French, Italian, Romanian and Ukrainian while keeping sensitive data within Swiss borders, a multilingual public-sector translation approach that could inform wider cantonal adoption.
Chacao municipality — Venezuela's Chacao municipality, Miranda state, runs AI robot dog patrols with two self-guided machines, Voltio and Turbo. Fitted with high-definition cameras and movement detectors, each streams live footage to the council's situational room, reads car number plates and spots stealing or vandalism of civic assets. Chacao calls this robotic patrol a Latin American municipal first.
Supreme Court of Montenegro and Judicial Council — Montenegro's Supreme Court and Judicial Council have rolled out an AI case law system built in-house by the Council's ICT department. It gives judges intelligent search across rulings plus an AI chatbot, unveiled at the Supreme Court on 8 April 2026, functioning as an AI-powered judicial search platform for the courts. A trial is under way at the Commercial Court, with a wider rollout planned across all Montenegrin courts.
Austrian federal government (State Secretary for Digitalisation) — Austria's State Secretary for Digitalisation, Alexander Pröll, says the federal electronic file system (ELAK) will add AI in the second half of 2026, letting 13,000 staff query whole file collections and get structured AI document summaries instantly. It is among five Public AI initiative applications, alongside KI-SUN, a knowledge assistant already covering 6,000-plus pages of federal training material.
Michigan Department of Health and Human Services — Michigan's Department of Health and Human Services now scans every SNAP case with an AI case-reading tool before payment, built on Google Vertex AI with optical character recognition reading pay stubs and other documents. David Knezek told the Senate Appropriations Subcommittee on DHHS that this AI benefits eligibility checking, an automated eligibility review, flags cases most likely to error, though staff still check each flag. It follows MiDAS, which wrongly charged over 40,000 people in the United States with misrepresentation within two years; a later audit of 22,000 cases found 93% involved no actual fraud.
Finanzamt Hilden — Employee tax cases in North Rhine-Westphalia now get AI tax risk management, with Finanzamt Hilden among the offices first to run it statewide in Germany. Risk-oriented selection flags audit-worthy returns, sorts low-risk filings, and clears some 2025 returns via AI-based tax audit selection with no human review. Eight offices piloted the approach from May 2025 before the wider launch.
Government of Bolivia — Bolivia's government has published an AI-generated government video reconstructing the police raid that took Uruguayan trafficker Sebastián Marset in Las Palmas, Santa Cruz de la Sierra. The animation follows officers into two luxury properties at around 3 AM on the Friday of the arrest, clearing the house holding Marset's armed guards first. Generative AI produced the sequence, out within hours of the raid.
National Information Technology Authority-Uganda (NITA-U) — Uganda's National Information Technology Authority (NITA-U) has rolled out UG Support, combining service requests, incident management, asset tracking, system monitoring and AI-assisted ICT service management in one interface. Proactive issue detection lets technical teams fix problems before users report them, linked into Uganda's e-citizen ecosystem via an AI-assisted support platform.
Togo Ministry of Economy and Strategic Oversight — Togo's economy minister Badanam Patoki unveiled the Digital Market Intelligence platform in Lomé on 12 March 2026, bringing AI fraud detection and data analysis to automated product-marking enforcement. SICPA's Togolese subsidiary built the tool, which flags potentially counterfeit consumer goods so Office Togolais des Recettes agents and Ministry of Commerce inspectors can aim inspections and market surveillance.
Government of New Caledonia — AI job matching arrived in New Caledonia through emploi.nc, a government portal pairing employers with jobseekers. The Government of New Caledonia built the site to generate structured CVs and deploy a matching engine that scans openings for the nearest fit to each candidate's profile, using occupational codes from France Travail.
Ministerio Público de Bolivia — Bolivia's Ministerio Público has run ARGO, the house-arrest arm of the ROMA–ARGO platform, since 10 March 2026, and AI house-arrest monitoring now reaches 275 people via a digital geofence, SkyFace facial recognition prompting random check-ins, and SONUS voice matching against the identity a judge authorised. Armando Corral Davezies, the Ministerio Público's technology director, puts the tally past 44,133 biometric check-ins on a platform needing no standing human oversight. In a Cochabamba field test the biometric supervision system caught a breach, georeferenced it and raised alerts inside three minutes.
Akwa Ibom State Judiciary — Among six states trialling courtroom AI for the National Judicial Institute, Akwa Ibom State Judiciary now applies AI in court proceedings, recording and converting hearings into written form without manual effort. LexScript AI built the underlying tool, delivering finished transcripts inside 24 hours. Justice Ekaete Fabian-Obot, the Chief Judge, foresees swifter case handling as this courtroom transcription pilot spreads further.
Government of Bermuda — The Government of Bermuda will pilot an AI pension chatbot so staff answer pension queries faster and more accurately, with the Swiss-based non-profit Near Foundation as partner. Success would widen the AI-powered pension tool into a web tax portal plus calculator, letting Bermudians inspect records and transact with government systems. A government spokeswoman set out the plan in a press release issued after Premier David Burt attended NearCon in San Francisco on February 23 and 24.
Nigeria Immigration Service — Ten years of traveller records are now searchable via AI visa overstayer tracking, thanks to the Nigeria Immigration Service's new Integrated Operations Centre. Minister Olubunmi Tunji-Ojo said the AI-driven immigration records system grants officers live history lookups, letting NIS catch overstaying foreigners nationwide amid wider border reform.
Government of Luxembourg — Luxembourg's government, launching the AI4LUX campaign, said every civil servant would get a sovereign AI chatbot within weeks, hosted on site under its Mistral AI partnership so data stays in the country. Staff are to process sensitive information securely with it and create personal agents for routine work. A legal chatbot for the Legilux platform is in development, and one for the most-visited state websites is being prepared.
Uganda Judiciary — An AI judgment writing tool now belongs to Uganda's Judiciary, co-created with the Lexis-Nexis Rule of Law Foundation. Slotted into Microsoft Word, it stocks templates, statutes and prior rulings, trimming research hours so decisions get drafted faster. Judge Geoffrey Kiryabwire dreamed up the concept and flags more AI-assisted judicial decision drafting capability on the way.
Irembo — An AI chatbot government services tool now answers Rwandans in Kinyarwanda, English or French on the IremboGov portal, live and fielding queries. The launch follows the platform's GovTech Prize win for Best Government Service in the World at the World Governments Summit in Dubai, and engineers are building a voice-based AI virtual assistant for citizen support next.
Hutt City Council — Hutt City Council's AI-Volution programme has put 15 custom-built AI assistants and 300 licences into staff hands, an example of AI in local government. The rollout frees 44,000 working hours a year and saves NZD 900,000 annually, with 26 staff certified as AI navigators. Licence holders complete compulsory e-learning first, aimed at faster, more responsive resident services through generative AI adoption among council staff.
Nepal Department of Customs — Nepal Customs applies AI customs risk scoring at six major offices handling nearly 90% of commercial consignments, through its machine learning Automated Customs Risk Management System (CRMS). Misclassification surfaced in 0.02% of declarations within three months, yielding about 3.74% extra revenue on them. Its weighted-parameter risk assessment profiles transactions and sends officers recommendations through ASYCUDA World.
Public Administration and Civil Service Bureau (SAFP) of Macao — Macao's Public Administration and Civil Service Bureau (SAFP) has introduced an AI platform for government departments, running large-scale reasoning, multimodal and language models from within the government cloud computing centre for exclusive internal office use. By the count, 28 departments and around 5,000 registered users have signed on, with the system handling tasks from document drafting and image recognition to content production, voice transcription and data analysis. The bureau has also added an AI office platform for public administration on its service hotline, helping staff handle public enquiries.
Nepal Police — Nepal Police opened the Artificial Intelligence and Advanced Analytics Cell (AI-AAC) at Kathmandu headquarters, applying AI in policing. Misinformation spreaders in five districts already face legal action through it. Machine learning, predictive analytics, generative AI and large language models power this police AI for cybercrime detection, early warning systems, proactive intelligence and countering disinformation.
Constitutional Court of the Czech Republic — Czechia's Constitutional Court has switched on a rebuilt virtual assistant for AI case-law search on the court's pages and in NALUS, after weeks of testing. Liberec's Technical University spin-off nanotrix.ai wrote the system, described by the Court as the earliest such tool on any Czech court site or judgment archive. The judicial database chatbot fields spoken queries and queries in other languages; users must verify whatever comes back.
High Commission of Rwanda in the United Kingdom — Rwanda's High Commission in the United Kingdom has launched Nshuti, an AI consular chatbot on WhatsApp and the mission's website giving the diaspora and others seeking services instant consular answers. The in-house assistant guides users on appointments, passport applications and renewals, document registration and certification, and national ID enquiries.
Nemzeti Adó- és Vámhivatal (NAV) — Hungary's Nemzeti Adó- és Vámhivatal now grounds its 2026 audit programme in AI tax audit selection. Vágujhelyi Ferenc, who leads NAV, sketched a single sweep across invoice feeds, till records, eKÁER consignment filings, sectoral declarations and cross-border money-flow reporting. Risky traders surface the instant a bill is cut, letting AI-driven risk screening put a reviser onto a fraudster ahead of any filing.
Belgeodesiya — Belgeodesiya is putting AI building recognition into the digital foundation of the State Navigation Map of the Republic of Belarus, per the State Property Committee. Accuracy runs from 65 to 90 per cent, varying with how many storeys a structure has. Settlements across 40 administrative districts have gone through satellite-image building detection, thickening the map's Buildings layer.
General Directorate of Taxes (DGI, Morocco) — Morocco's General Directorate of Taxes announced on 18 February an AI property valuation trial using public property-price comparables in Rabat and Casablanca, built with Moroccan start-up Yakeey. Director general Younes Idrissi Kaitouni reports approval and signed contracts, and the price comparables should end valuations varying by inspector in property and transfer-pricing audits.
Tanzania Judiciary — Speaking a ruling aloud now suffices: Tanzania's judiciary's AI transcription for courts converts the spoken words into finished text, rendering it between Kiswahili and English in minutes instead of weeks. Italian firm Almawave built the tool, tested since 2024 across 11 judicial zones under the Transcription and Translation Software programme, and courtrooms beyond that pilot are next. Checking documents is also getting quicker as this AI-powered court translation software links data with the Tanzania Revenue Authority and the Business Registrations and Licensing Agency.
Judiciary of Iran, Statistics and Information Technology Centre — Iran's judiciary unveiled 'Smart SAMP', an AI case management system first piloted in three provinces, South Khorasan, Qom and Lorestan, from Dey 1404 (late December 2025 to January 2026). Judges each get an assistant that answers legal queries using legal sources and harmonises precedents and advisory opinions, while smart tools handle case referral and expert-review assignment, says the judiciary's statistics and IT centre head.
Central Agency for Information Technology (Kuwait) — Kuwait's Central Agency for Information Technology will set up, with Sprinklr, one system for handling citizen and resident enquiries across every state body, starting with an AI contact centre overhaul of Wasel 101. The agency says Wasel 101 will then sort and forward queries automatically, run real-time analytics and give the AI-assisted call centres virtual assistants and smart agents.
Erbil Governorate — Erbil Governorate flipped the switch on Project Judy on 12 February 2026, hailed as the first AI legal research system anywhere in Iraq or the Kurdistan Region. The launch at the Kurdistan Bar Association drew the Governor and the Minister of Justice, with courts, lawyers and public agencies expected to lean on the new tool for case analysis, evidence discovery and file handling. Organisers say the AI-driven legal assistant should thin out backlogs and take weight off judges.
Libyan Audit Bureau — Libya's Audit Bureau has rolled out an AI assistant for government audit staff called DewanBot, built in-house by its own IT office. The tool helps employees draft official reports and correspondence while cutting the time needed to locate information, forming part of a wider digital transformation push within the Bureau. This AI-assisted auditing deployment reflects a broader adoption of modern government technology tools.
Prosecutor General's Office of Azerbaijan — Azerbaijan's Prosecutor General Kamran Aliyev confirmed automated triage for AI in public prosecution: since late 2025, systems suggest which officer handles each application, staff finalising it. The effort extends to AI government case management, with tools planned for investigations and citizen phone enquiries under supervision.
Central Elections Commission of Palestine — The Palestinian Central Elections Commission built Nazih, an AI voter assistant for the April 2026 local elections and new election law. The AI-powered election chatbot runs on the commission's website, WhatsApp and Facebook, logging 22,000 queries in its January 2026 trial.
Superintendencia de Administración Tributaria (SAT) — Guatemala's SAT will introduce AI tax filing through its FEL 3.0 electronic invoicing platform, generating pre-filled ISR (income tax) and IVA (value-added tax) declarations from 2026. SAT chief Werner Ovalle said the system enables risk management and automated cross-checking, building on FEL 2.0 analytics that already caught invoice splitting and anomalies. This AI-generated tax declarations layer extends existing detection tools rather than replacing them.
Gobernación del Atlántico — Atlántico's departmental government awarded its Intelligent Emergency and Security System (SIES) to put 414 AI video surveillance cameras, with facial recognition and automatic number-plate reading, in its municipalities. The Sabanalarga police command will watch the smart cameras centrally, piloting first at Santo Tomás's Batalla de Flores on 7 February.
General Traffic Department, Ministry of Interior (Kuwait) — Kuwait's General Traffic Department will soon bring AI traffic enforcement to delivery vehicles, public transport, car rentals, ride-hailing services and taxis through an electronic monitoring system. The system will watch driving behaviour, spot violations and issue automated citations in real time, supporting prosecution of offenders.
Ghana Revenue Authority — The Ghana Revenue Authority says customs takings rose from GH¢2.4 billion to GH¢3.6 billion within weeks of Publican AI going live at Tema Port. This AI customs valuation system checks declared import prices against global trade databases with machine-learning models and routes suspicious entries to a customs officer before clearance. Freight forwarders and importer groups call it opaque and slow at the busiest port, prompting an April media briefing.
Dirección del Parque Nacional Galápagos — Floreana's Smart Island system in the Galápagos has fed real-time AI wildlife monitoring to park rangers, researchers and technicians since 1 February 2026. Trained on thousands of photos from 125 camera traps and 48 wireless cameras, it recognises feral cats and rodents. The Galápagos National Park Directorate, which leads Floreana's restoration alongside the Jocotoco Foundation and Island Conservation, relies on it to pick and monitor habitats for returning species like giant tortoises.
Vietnam Customs (Ministry of Finance) — A trial AI chatbot for customs clearance runs in Vietnam at chatbot.customs.gov.vn until 30 April 2026. Businesses and officials use it for HS code lookup, duty rate suggestions and trade queries.
HM Government of Gibraltar — Gibraltar's Central Arrears Unit now runs on a bespoke AI debt collection database, built after a hiring round for several vacancies yielded a single suitable applicant. Nigel Feetham, Minister for Justice, Trade and Industry, explained the gap; IT Development Manager Karon Cano wrote the model in-house with Steven Fernandez. Machine-assisted arrears tracking replaces hand-keyed spreadsheet extracts and saves one officer at least an hour's work. A chatbot with no outside feeds returns priority-debt listings.
Alcaldía Mayor de Bogotá — Bogotá's city government has upgraded Chatico, its generative AI citizen assistant, to answer residents and visitors on public services and municipal procedures more precisely and fluently. The virtual agent runs under the Bogotá Capital Digital strategy, managed by the District ICT Office and fed by several district agencies.
General Authority of Customs (Qatar) — Qatar's General Authority of Customs launched AI customs classification and other AI services on 26 January 2026, opened by the finance minister on International Customs Day. Smart Harmonized System classification proposes the tariff code from the first shipment-data entry, pre-classification hints speed and sharpen customs officers' decisions, and an agentic AI service and customs-information virtual assistant join them.
General Audit Chamber of St. Maarten — St. Maarten's General Audit Chamber launched Audit Intelligence, a GPT giving citizens natural-language access to verified public-sector data, advancing AI-driven government transparency. It covers budgeting, spending and oversight, sourced from the Audit Chamber, central bank, statistics bureau, World Bank and IMF: AI-powered public sector oversight in practice.
Department of Foreign Affairs and Trade — Ireland's Department of Foreign Affairs and Trade is piloting Ask Finn, an internal AI chatbot for passport processing, as Passport Service, Revenue and Citizens Information Board officials prepared for the Oireachtas Committee on Artificial Intelligence. Revenue's AI models route taxpayer queries more precisely to caseworkers, and Citizens Information is rolling out its AskCI bot.
Consell Superior de la Justícia — Andorra's Consell Superior de la Justícia has switched on an AI case-law search holding 15,000-plus rulings, open to lawyers and citizens. Machine-generated metadata tags each entry, letting queries run in plain speech or boolean operators across every decision under one statute. A 4,000-term thesaurus guides the rulings archive, with ECLI codes for cross-border traceability.
Servicio Nacional de Aduanas de Chile — Chile's customs service has started operating an AI cargo scanning truck at the Iquique International Terminal in Tarapacá. The scanner truck is meant to bolster regional checks against drug trafficking and smuggling.
Ministry of Industry and New Technologies of Tajikistan — Tajikistan's Ministry of Industry and New Technologies has launched AI-GOV, deploying the Perplexity Comet browser across government offices as a step toward AI in public sector administration. Epsilon3.ai built the system architecture as technology partner, and the tool already runs inside the Ministry, where it handles licensing approvals for small and medium businesses, covering regulatory checks and knowledge management, an early government use of an AI browser.
Dubai Municipality — Dubai Municipality is piloting AI litter detection with cameras on waste trucks watching roads, neighbourhoods and public spaces for littering and illegal dumping. Instant image analysis puts violations on digital dashboards for quick field action; littering costs up to Dh500, under the emirate's Smart Waste Management framework.
Poder Judicial del Perú — Peru's Poder Judicial has put ChatBot CURIA, an AI case-law assistant, into the National Superior Court of Specialised Criminal Justice, the Superior Court of Justice of Lima and the Superior Court of Justice of Lima Norte. Built in-house by the Supreme Court's Fifth Transitional Chamber, the intelligent search tool covers over 500,000 fully integrated rulings, may reach 6 million jurisprudential records nationally, and answers judges and staff on legislation, case law and doctrine. Judge Ulises Yaya Zumaeta says the final decision-maker must corroborate every AI answer.
Council of Ministers Secretariat General (Qatar) — Qatar's Council of Ministers Secretariat General opened the first phase of its 'Smart Legislative Advisor', an AI legislative drafting tool built with the Ministry of Communications and Information Technology. The legal-drafting AI checks drafts against regional and international laws, tests consistency with the constitution and current legislation, and handles language and proofreading review, supporting legislative professionals rather than replacing them.
General Department of Taxation (GDT) — Cambodia's General Department of Taxation has deployed Beauramei, an AI tax chatbot answering queries on tax laws in Khmer, cutting manual enquiries and office visits. GDT ties the rollout to a wider modernisation drive for public services and stronger domestic revenue mobilisation, with further AI-powered tax services upgrades to follow.
Ministry of Posts and Telecommunications — A free AI translation app, TranslateKH 2.0, went live from Cambodia's Ministry of Posts and Telecommunications, letting anyone switch text between Khmer and English via the Play Store, the App Store or a fresh site, www.translate.kh. Powered by deep learning and a large language model, the second-generation machine translation service reads whole passages for meaning instead of parsing sentence by sentence, which the ministry credits with sharper results than its predecessor.
US DoD Defense Innovation Unit — The US Department of Defense's Defense Innovation Unit began prototype testing of the Joint Sustainment Decision Tool, AI for predictive military logistics, in August 2026, with tests running to March 2027. It is meant to foresee and reduce risks across military sustainment and the supply chain.
The UK Home Office launched PoliceAI, a national centre for AI in policing, on 10 June 2026 with £75 million over three years to find, test and scale AI tools across the 43 forces of England and Wales. Named uses: facial recognition, deepfake detection, call transcription, triage of non-emergency calls and digital evidence, disclosure and audio-visual redaction. It says the tools could eventually free the equivalent of 3,000 officers.
The Earth Fire Alliance, with CAL FIRE, launched the first three operational FireSat satellites, billed as the first system built to detect wildfires, from Vandenberg Space Force Base on 7 July 2026. Three months of commissioning and calibration precede operational data for fire agencies. The full constellation should exceed 50 satellites by 2030, covering the planet every 20 minutes; CAL FIRE did early testing and concept evaluation.
Primăria Sectorului 3, Bucharest — Romania's Primăria Sectorului 3, Bucharest's Sector 3 city hall, now answers residents' calls with an AI phone assistant handling routine administrative queries instead of a human operator. Reported on 21 April 2026 after weeks of use, the intelligent virtual assistant was still in calibration; the city hall admits errors can occur meanwhile.
Israel Tax Authority — Israel Tax Authority Director Shay Aharonovich says AI tax audit tools have uncovered NIS 20 billion in discrepancies, with AI-based checks comparing capital declarations against the authority's other financial data. At the Eli Hurvitz Conference on Economy and Society in Jerusalem, he said about 4 per cent of filings are scanned today.
Commonwealth of Massachusetts — Massachusetts governor Maura Healey announced that the state's 40,000 executive-branch employees will get a ChatGPT-powered AI assistant, deployed with OpenAI after competitive procurement. Rollout is phased, beginning with the Executive Office of Technology Services and Security, then other agencies, and the state says the tool keeps prompts in a walled-off environment where they cannot train public models. Other states have run limited ChatGPT pilots; this is the first across a whole branch of state government.
Los Angeles County Superior Court — Los Angeles County Superior Court is piloting Learned Hand, an AI tool that works as a judges' clerk, distilling long filings, writing research memos and drafting orders. The court's contract, worth roughly $314,000, maps out testing in criminal, family and probate divisions, though judges so far mostly use it on civil motions; Riverside County's court is trialling it under a $10,000 agreement. Learned Hand combines Anthropic, OpenAI and Google models, and says it checks bias and accuracy but has published no results.
Judiciary of Kenya — Kenya's Judiciary has moved from trial to routine AI court transcription, capturing proceedings electronically with transcripts ready overnight. Supreme Court Judge Isaac Lenaola presented it at the AI Everything Summit at Gitex Kenya in Nairobi, adding that the Judiciary plans a countrywide hub to pool transcription for all courts. A Judiciary chatbot also responds to public legal queries.
Tax Committee of Uzbekistan — The Tax Committee of Uzbekistan has launched AI SOLIQ, a single AI tax platform running on its proprietary tax database. Entrepreneurs, accountants and company directors can chat to check their organisation's tax rating, available reliefs and filing duties, while tax officials apply it to decision support, routine automation and data analysis. It runs on the committee's Soliq LLM with machine-learning, computer vision and speech generation components.
Shizuoka City — Shizuoka City has opened a generative AI enquiry system to all staff on Allganize's Alli LLM App Market, answering officers' questions from internal manuals and working documents. Each answer shows the passage it relies on so officers can verify the basis before acting, and the city hopes to shift time saved on internal queries to resident services and policy work.
Western Australia Police — Western Australia Police have scanned over 130,000 people in Perth and Fremantle since a live facial recognition trial began in June 2026, using NEC's Neoface m40 software from a marked van at major events and crowded places. Real-time checks compare faces against a watchlist of missing people, registered child sex offenders and people with outstanding arrest warrants, and matches alert nearby officers.
Supreme Court of Kazakhstan — The Supreme Court of Kazakhstan is finishing the nationwide rollout of its Digital Judge's Assistant, an AI judicial assistant piloted for about eighteen months in Astana and the Zhambyl, Karaganda and Mangystau regions. It searches about 15 million machine-readable units of judicial material for comparable cases, analyses practice, structures claims, lays out parties' positions in Kazakh and Russian and forecasts likely civil-dispute outcomes. Judges keep the decision, with court chairs stressing the tools assist rather than replace them.
Dhaka Metropolitan Police — Dhaka police linked city traffic cameras to AI traffic enforcement software in April 2026 for automatic violation detection, from illegal parking to lane and signal breaches. Officers check each case before a fine, owners get a text message, and at least 300 vehicles have been prosecuted, with serious offenders fined and others warned. Manual enforcement will gradually end, police said.
Ministry of Economy and Development of Mongolia — Mongolia's Ministry of Economy and Development has built AI Туслах, an AI planning assistant, into the government's digital planning platform guiding officials live on data-driven KPIs. With the Partnership for Action on Green Economy, the ministry aims it at planners in ministries and agencies who set measurable indicators and pull together national development plans, especially newcomers and staff in high-turnover posts. It remains a pilot.
National Tax Service of Korea — Korea's National Tax Service has offered taxpayers a generative AI tax chatbot since January 2026, first for VAT returns and year-end tax settlement, widening it from 1 May to comprehensive income tax filing and incentive claims. It sits beside an AI telephone consultation line and AI search in the Hometax portal, public-facing first. The NTS plans to extend AI from taxpayer services to tax-base management and audits across all tax administration.
Varsinais-Suomi and Satakunta wellbeing services counties — The Varsinais-Suomi and Satakunta wellbeing services counties launched joint work in June 2026 to take AI record-keeping in social care from pilots to full social welfare services. Over 600 Varsinais-Suomi professionals created 30,000-plus AI-assisted records in the pilot, about 300 in Satakunta produced over 10,000 in 2026, and a tenfold rise in use is targeted by end-2027.
Национална агенция за приходите (NRA) — Bulgaria's National Revenue Agency is extending AI in tax administration to all its units after a successful pilot, the first Bulgarian administration to use the national BgGPT model at scale. Built with INSAIT, it covers head office and seven territorial directorates, targeting everyday use by 7,000-plus staff searching specific cases under shifting tax and social-security law.
Dutch central government (Ministry of the Interior and Kingdom Relations) — The Dutch government is extending vlam-chat, a home-built AI chatbot for civil servants, to several units of the Ministry of the Interior and Kingdom Relations, with pilots elsewhere in central government. The ministries of Health, Finance, General Affairs, Education, and Infrastructure and Water Management, plus the High Councils of State, already have it. 82% of surveyed users use it at least weekly and 77% save noticeable time; production rollout follows in the second half of 2026.
North Rhine-Westphalia tax administration (Finanzverwaltung NRW) — North Rhine-Westphalia's tax administration is live-testing FinAssist.NRW, an in-house AI assistant hosted in its own data centre. Tax offices in Gütersloh, Oberhausen-Nord and Solingen have trialled the chatbot live since June 2026, before the state finance ministry's own trial; it condenses and analyses internal papers, writes drafts, organises email chains and pulls out task lists.
Government of Côte d'Ivoire — The Ivorian government has launched EMY 101, an AI government chatbot on WhatsApp and Messenger that explains state policy and official formalities, from civil registration to tax. Users can also voice worries, flag alerts and find the correct public-service contact through the citizen assistant, supported by the free 101 line.
Agencia de Recaudación de la Provincia de Buenos Aires (ARBA) — Buenos Aires province's tax agency ARBA launched M2, AI property tax detection combining high-resolution satellite imagery, drone surveys and AI models to find undeclared buildings and improvements. Detection, analysis, notification and the taxpayer's reply share one digital environment, and ARBA says the automated monitoring greatly widens its capacity to regularise such cases.
General Directorate of National Security (DGSN, Morocco) — Morocco's national police (DGSN) have deployed two home-designed AI patrol vehicles, Amane and Madar, in Rabat, with Casablanca and Tangier next before wider rollout. The smart patrol cars' 360-degree cameras feed real-time facial recognition and automatic number-plate reading through AI algorithms for instant checks; officers trained first, the DGSN says.
Ministry of Treasury and Finance (Türkiye) — Türkiye's Ministry of Treasury and Finance has deployed KAŞİF, AI fake invoice detection that screens about 4 million taxpayers daily. Built by the Revenue Administration's GİB Teknoloji and the TÜBİTAK Artificial Intelligence Institute, the machine-learning system picks out high-risk taxpayers likely to issue fake documents for Tax Inspection Board audits.
Federal Court of Appeal — Canada's Federal Court of Appeal built Amicus, an AI assistant for court registry staff answering procedural and operational questions. It draws sourced, verified material from a central database, so matching queries yield matching answers rather than results that vary by employee. Feedback has been strongly favourable; the goal is consistent, national service quality for internal teams and external litigants alike, functioning as a broader AI registry assistant.
City of Burlington — eCheck lets applicants for a Burlington building permit pre-test designs free before formal submission. The Canadian City's AI building permit review tool checks proposals against zoning bylaws and the Ontario Building Code, covering ceiling heights and room sizes plus fire protection, setbacks, height, floor ratios, landscaping and parking. Mayor Marianne Meed Ward said Burlington is first in Canada to apply automated building code assessment pre-submission, ahead of other municipalities on early compliance checks.
Louisville Metro Government — Louisville Metro Government's new AI housing permit process pilot, starting this January, picks Govstream.ai to scan zoning codes, geographic records and past filings for recurring delays. Officials expect faster reviews, clearer visibility and earlier guidance for applicants and staff alike. Outcomes from the AI-driven building permit scheme go public once the spring trial ends.
Royal Gibraltar Police — The Government of Gibraltar has brought Live Facial Recognition into round-the-clock service along the reopened frontier and further priority sites, with the Royal Gibraltar Police as data controller. A heavy volume of stills ran through AI face matching in week one, flagging nobody on a police watchlist. Recordings are kept to prevent and detect crime and protect public safety and national security. Chief Minister Fabian Picardo rates the technology estate, facial recognition included, as holding up well.
Sociaal Economische Raad (SER) Curaçao — The Social and Economic Council (SER) of Curaçao keeps generative AI in advisory work under safeguards fixed at master classes on March 19, 20 and 24 at the council's Ansinghstraat offices. Condensing long files, sorting material, polishing wording, matching documents and simplifying dense subjects go to the software; legal interpretation, weighing competing interests, institutional sensitivity and formal judgment do not. Three rules govern AI-assisted advice: humans answer finally, personal data is protected, and quality, integrity and verifiability hold. Director Raúl Henriquez: nothing displaces professional judgment or human responsibility.
Nagyoroszi Council — Nagyoroszi's council in Hungary has opened a freely reachable AI customer service, supplied by Munipolis, still learning and fed solely by the village's own web pages. Preset bubbles cover the commonest queries, typed messages work too, and adjacent topics appear unprompted. The municipal chatbot copes with several languages, spots the edge of what it knows and hands anything past that to whichever council officer owns the case; officials report lighter paperwork already.
Government of Iceland (Ministry of Finance and Economic Affairs) — Microsoft Copilot is spreading through Icelandic state institutions under a three-year Microsoft agreement with the Ministry of Finance and Economic Affairs, driving a coordinated push for AI in government services. Officials cite better public services, stronger staff, greater efficiency and responsible artificial intelligence in the public sector as goals. Rollout speed depends on each institution, with large-scale training starting autumn 2026.
Direction Régionale des Finances Publiques (DRFiP) de La Réunion — Direction Régionale des Finances Publiques (DRFiP) Réunion applies AI to aerial photographs, uncovering over a thousand unreported swimming pools for property tax purposes. This AI tax fraud detection effort, running since 2025, is expanding to catch unpermitted building work like garages and terraces, with AI-assisted tax compliance also helping agents verify claimed expenses during income tax filing periods.
Tax Committee of Tajikistan — Tajikistan's Tax Committee runs AI in corporate tax administration, matching fresh registrations against state records and confirming entrepreneur identities and paperwork through the Andozi Man mobile app. AI property tax algorithms combine several government data sources to assess property tax, while a separate electronic tool handles public tax queries.
Government of Malta — Malta's Government wants machines handling court sittings, parliamentary sessions and public submissions: a tendered AI speech-to-text system, priced near €147,750, targets 85% accuracy in year one while managing code-switching between Maltese and English. Staff will check and fix computer drafts via a smart editor — an automatic transcription tool for courts and parliament.
Government of Burkina Faso (via CITADEL) — CITADEL researchers built the AI voice translation Mooré tool now live on Burkina-en-marche.gov.bf, delivering spoken health and education content rather than written text. Converted French speech hits a 4.24% word error rate, a standard 77.8% of tested native speakers accepted. This French-to-Mooré speech translation reaches those who never learned letters.
Central Bank of Liberia — Helen Momo built the Central Bank of Liberia's new AI macroeconomic forecasting system, calling it the region's "First Sovereign AI Infrastructure for Predicting Macroeconomic Forecasting and Financial Intelligence in the Region." Shown publicly at the Liberia Technology Summit, the tool generates financial intelligence and projects economic trends for the bank. Momo, a Harvard-educated technologist from Liberia, designed it to serve the institution directly, giving the country a homegrown alternative for economic trend prediction rather than relying on outside systems.
Turks and Caicos Islands Government — The Turks and Caicos Islands Government intends AI parliamentary transcription via an AI-assisted Hansard on a new legislature portal, turning spoken sittings into indexed text. Minister of Innovation, Technology and Energy E. Jay Saunders targets the 50th anniversary of Ministerial Government. Lawmakers and citizens gain easier lookup through machine-generated House minutes.
Dewan Ekonomi Nasional (DEN) — A national AI government procurement platform is being readied in Indonesia for rollout between October and November 2026, National Economic Council (Dewan Ekonomi Nasional) chairman Luhut Binsar Pandjaitan revealed, with unveiling expected from President Prabowo. Pricing irregularities in state purchasing would be flagged automatically, trimming spending by 20 to 30 per cent, while facial recognition paired with integrated databases pushes social aid targeting accuracy past 90 per cent. Corruption should fall too in this shift toward AI in public administration, since the design strips human involvement out of administrative workflows.
Chacao municipality (Alcaldía de Chacao) — Turbo and Voltio, two AI-powered robot dogs public safety patrols units, roam Chacao's squares in eastern Caracas after dark. Cameras and motion sensors relay live video to the command centre, helping staff spot odd movement and verify plates. Gustavo Duque called the AI-powered robotic patrol units support, not a stand-in, for police.
Somali Disaster Management Agency (SODMA) — Meteorological drought conditions across Somalia can now be flagged three months ahead of rainy seasons, thanks to a machine learning drought prediction tool handed to Somalia's Disaster Management Agency and Somaliland's disaster authority. FutureWater and Galayr built the drought forecasting model under a project commissioned by the World Food Programme, shifting response away from reactive crisis handling toward anticipatory action.
Suseong-gu Office, Daegu — Suseong-gu in Daegu is trialling an AI chatbot for residents through October, covering administrative lookups and facility bookings. Multilingual, voice-enabled support aids foreigners and digitally excluded people; full AI civil-service help starts in November.
Polish Ministry of Digital Affairs (Ministerstwo Cyfryzacji) — Citizens across Poland can now ask an AI virtual assistant for government services directly inside the mObywatel app, following its 31 December 2025 debut from the Ministry of Digital Affairs. The tool runs on PLLuM, a homegrown Polish Large Language Model, and works as an AI chatbot for public services, walking people through tasks such as requesting a passport or ID card, securing a PESEL number, or notifying authorities of a car sale. It cannot see personal records or tap into state databases, and only reached the public after a two-month trial limited to 5% of users.
State Tax Service of Kyrgyzstan — A newly deployed system called Kezet now underpins AI tax auditing at Kyrgyzstan's State Tax Service, built by the Salyk Service state institution to strengthen how data gets processed. Inspectors will lean further on AI-driven tax inspections from 2026, when audits themselves become automated and oversight grows more transparent.
Mongolian Customs General Administration — MCGA now leans on AI HS Code classification to speed cargo clearance in Mongolia. A chatbot combining a large language model with retrieval-augmented generation runs atop Rebellions NPU chips, so traders keying in goods descriptions instantly get back a suggested harmonised tariff code, trimming errors and wait times at the border. Mongolian Customs General Administration bills this as the world's first NPU-based LLM/RAG setup embedded in a working customs system, an early real-world proof of AI tariff coding for customs.
Public Prosecution (Oman) — More than 90 percent accuracy: that's the figure Oman's Public Prosecution cites for a fresh AI judicial document analysis trial developed alongside Rahal Company. Rather than moving case files elsewhere, the tool composes questions for parties involved, readies rulings destined for storage or onward referral, and holds sensitive material on in-house systems throughout. It forms one strand within a wider experimental push from the Ministry of Transport, Communications and Information Technology, covering health, procurement, cybersecurity, open data and education alongside AI in the judiciary.
Anguilla House of Assembly — Anguilla's House of Assembly deployed parliamentary AI transcription via its Hansard AI Project, transforming sitting audio into searchable, downloadable PDFs matching classic Hansard formatting. A chat-box built into the AI Hansard system fields queries on Session Bills and members. The capital budget covers costs.
City of Windsor — Windsor's Provincial Offences Administration office will launch a voice-enabled AI chatbot for municipal government, answering resident calls over a six-month trial. Officials forecast it absorbing about 30 per cent of clerks' duties, with its $57,600 yearly cost offset in 10 months via staffing savings from this virtual assistant deployment.
Dún Laoghaire-Rathdown County Council — Gaist's AI road condition survey system now covers 823 km of routes for Dún Laoghaire-Rathdown County Council, Ireland's first such pilot by a local authority. Vehicle cameras capture 360-degree footage that AI software turns into surface ratings, faded-marking alerts and street furniture maps, feeding wider AI-based infrastructure monitoring.
Arbeidstilsynet (Norwegian Labour Inspection Authority) — Arbeidstilsynet, Norway's Labour Inspection Authority, has launched an AI chatbot for foreign workers built on material drawn from arbeidstilsynet.no. The AI-powered assistant spans the regulator's entire area of responsibility and replies in 12 languages: Norwegian, English, Bulgarian, Estonian, French, Latvian, Lithuanian, Polish, Portuguese, Romanian, Russian, Spanish and Ukrainian, helping migrant employees get guidance in their own tongue. Its debut supports a broader state campaign targeting social dumping and workplace crime.
Federal Civil Service of Nigeria — Nigeria's Federal Civil Service has logged 25,000-plus chats on Service-Wide GPT, an AI tool for public administration. It draws on the Constitution, Public Service Rules, statutory regulations and official circulars for workers across ministries, departments and agencies. Didi Esther Walson-Jack cited the count as evidence the AI chatbot for government staff is catching on.
Basic Court in Zrenjanin (Osnovni sud u Zrenjaninu) — Instant written transcripts now emerge from Zrenjanin's Basic Court hearings through Jurisdikta, an AI speech-to-text court system offering speech recognition for legal proceedings, confirmed live by the presiding judge. Serbia's judicial officials additionally rolled out Transkripta, turning 800 hours of recorded audio into text inside a single 24-hour span, alongside Legislativa, a chatbot pulling answers solely from domestic statutes, regulations and precedent rather than the broader internet. A gathering on Kopaonik mountain hosted the debut of all three systems.
Ministry of Internal Affairs of Azerbaijan — Azerbaijan's Ministry of Internal Affairs added an AI chatbot for police services to the E-Polis mobile app and the online portal e-xidmet.mia.gov.az, starting as a pilot. Users type questions and get immediate automated replies covering applications, required documents and procedures. Pending trial outcomes, the policing assistant will gain voice features and sharper replies.
ICT Division, Government of Bangladesh — kagoj.ai, Bangladesh's first Bangla-language AI platform, has gone live from the ICT Division alongside 'July', a new government typeface. Nearly 4,000 volunteers trialled it over a fortnight before official debut. Authorities plan to release the API and code openly, while extending efforts to gather spoken-word data from ethnic minority communities using this AI platform built for Bangla.
Hamburg Senate Chancellery — Roughly 40,000 civil servants now use LLMoin, the City of Hamburg's GDPR-compliant AI writing assistant for government use, after a phased rollout. Dataport built it to draft, summarise and search documents. Bremen, Niedersachsen, Sachsen-Anhalt and Rheinland-Pfalz separately adopted the same AI text drafting tool for civil servants.
Instituto de Gestão Financeira e Equipamentos da Justiça (IGFEJ) — IGFEJ (Institute for Financial Management and Justice Equipment) switched on AI call routing for court services within Portugal's justice administration. An audio bot answers callers, works out their needs, and forwards requests automatically instead of a human operator. Coverage at start spans costs, registry, judicial matters and the technology catalogue.
Procuraduría General de la República (PGR) — El Salvador's Procuraduría General de la República introduced an AI virtual agent for public defenders on 11 December 2025, for staff handling child and adolescent cases. Running on the agency's own secure infrastructure, it merges national rules and case criteria to verify requirements and draft legal inputs faster, using artificial intelligence for legal support as part of the Procuraduría's modernisation drive.
Ministry of Communications and Information Technology and National Planning Council — Scale AI is joining Qatar's Ministry of Communications and Information Technology and the National Planning Council to craft an AI-powered decision-making tool for public bodies. Officials hope it will forecast trends and steer policy using economic and social data, reflecting a wider shift toward AI in government data analytics.
Department for Work and Pensions — The Department for Work and Pensions now applies AI vulnerability detection in welfare correspondence, running its Whitemail Insights and Vulnerability Scanner across roughly 25,000 letters and scanned documents daily. Built on an LLM, the tool spots warning signs spanning mental health, suicidal thoughts, hardship and abuse at home, generating AI-flagged vulnerable citizens for staff attention. Caseworkers still judge every flag.
Addis Ababa City Administration — Addis Ababa City Administration switched on the Integrated Civil Registration and Residency Service Technology, an AI-powered civil registration platform, across 50 woredas from 8 December 2025. The Ethiopian Artificial Intelligence Institute built this digitised civil registration system, linking vital events registration with residency identification for smartphone use. Mayor Adanech Abiebie called it a decisive move toward digitised services.
Servicio de Administración Tributaria — Fiscal risks now surface for Mexican taxpayers before they even file, under a new predictive audit pilot from Servicio de Administración Tributaria (SAT), the country's tax authority. Branded 'Auditorías Predictivas', this AI tax auditing effort draws on machine learning, big data and other artificial intelligence tools across upwards of 15 domestic and cross-border databases, spanning filings, CFDI digital invoices, banking activity, customs paperwork and social security records, to catch inconsistencies early and push warnings straight to taxpayers' Buzón Tributario inbox. Coverage so far centres on heavy CFDI issuers, online platform vendors, firms trading internationally and taxpayers with a risky compliance history — an early instance of AI-driven tax compliance monitoring in practice.
Ministry of Water Resources and Irrigation of Egypt — Over 55,000 kilometres of Egypt's canals and drains fall under a fresh AI water management push from the Ministry of Water Resources and Irrigation, part of the Second Generation Egyptian Irrigation System. Predictive software estimates coming Nile River levels to sharpen supply, while airborne and orbital sensors spot weed growth, catch rule-breaking and log riverbank changes. Hani Sewilam unveiled the effort at the World Water Congress in Morocco as AI-driven irrigation management nationwide.
Urząd Miasta Poznania (Poznań City Hall) — Poznań is piloting ShpaRAG, an AI assistant for citizen services built on Poland's PLLuM language model, developed with the HIVE AI consortium and NASK. It uses retrieval-augmented generation so call centre staff find answers to resident queries faster. As an AI-powered public service assistant, it joins other tools the city runs, including a generative AI helpline consultant and voicebots for appointment reminders.
Badan Pengawas Obat dan Makanan (BPOM) — Indonesia's Food and Drug Supervisory Agency (BPOM) now runs the nation's first AI drug and food licensing system, pairing automated workflows with machine learning and language-processing software wired into the National Data Centre to shrink approval turnaround. A sister tool, PRIMA, gives applicants a 24-hour AI-powered regulatory consultation portal for licensing enquiries.
Policía Nacional Civil (PNC) — Over 6,000 smartphones now carry El Salvador's ONI system, a mobile AI platform issued to National Civil Police officers and trainees under Plan Control Territorial Phase III. Acting as a modern police AI surveillance tool, it swaps out legacy radios for LTE-connected handsets that read identity papers and vehicle plates, check offender databases and single out suspected gang members, claiming above 40% precision. This AI policing platform's accuracy figure comes from the force itself and remains unverified.
Supreme Court of Oromia — Afaan Oromo court hearings will get an AI court transcription system under a new pact between Ethiopia's Supreme Court of Oromia and the Ethiopian Artificial Intelligence Institute (EAII), turning spoken arguments into written records automatically. Slow manual conversion of oral testimony had long frustrated the Oromia bench, prompting the collaboration. EAII's Dr Worku Gachena called natural language processing central to the institute's mission, framing this as an instance of automated speech-to-text for legal proceedings.
Berlin Senate Chancellery — Berlin's Senate Chancellery has equipped 40,000 civil servants with BärGPT, a government AI assistant for public administration offering drafting, translating, analysing and summarising via a secure chat interface. CityLAB Berlin built it as open-source software on GDPR-compliant servers in Germany, an AI tool for civil servants made locally, not licensed from a private vendor.
Ministry of Science and ICT / Ministry of Public Administration and Security — Hundreds of thousands of civil servants in South Korea can now tap private-sector AI tools inside government networks under a new AI common base service government employees jointly built by the Ministry of Science and ICT and the Ministry of Public Administration and Security. Samsung SDS and Naver Cloud built the first two chatbot tools running on this shared system, with further models set to join in the coming year. Officials say the arrangement will stop agencies from separately funding overlapping AI projects, folding access into a single intelligent work-management platform as part of broader AI adoption in public administration nationwide.
Public Service Commission of Tonga — Tonga's Public Service Commission piloted an AI chatbot for civil servants in November 2025 — an AI-powered guidance tool for government staff answering public service policy questions instantly. Harvard intern Molonai Hola built it with PSC input for ministry practices, unveiled at a briefing for HR and IT officers across 16 ministries. Wider adoption hinges on the trial's results, the commission notes.
Isle of Man Government (Cabinet Office) — 350-plus civil servants in Isle of Man's government now rely daily on Microsoft Copilot under an AI-driven government productivity scheme, the Efficiencies and Change Programme, targeting £50 million saved within five years. Cabinet Office confirmed by May 2025 departments were digitising workflows via AI tools boosting workforce output.
Kenya Revenue Authority — Kenya Revenue Authority is among Sub-Saharan Africa's earliest adopters of AI risk engines for tax compliance, built into its enforcement pipeline. Customs receipts hit a record KSh85 billion in one month once the pilot began. Fraudulent filings, mismatched income and risky transactions now get flagged by these machine-learning tools for tax enforcement.
Ministry of Finance / Lebanese Customs — Sixty containers an hour: that is the throughput now claimed for new AI cargo scanners customs teams are using at Beirut Port, under a wider reform drive in Lebanon. An AI programme reads each container's contents, keeps images on file for six months, and forwards scans to foreign customs agencies, Finance Minister Yassine Jaber said. Phase one of the plan covers this 60-container-per-hour equipment; later stages bring an integrated IT system and AI-based risk analysis.
Agenția Națională pentru Ocuparea Forței de Muncă (ANOFM) — Moldova's National Employment Agency (ANOFM) has fitted its angajat.md platform with an AI chatbot for job seekers, built with the Ministry of Labour and Social Protection and Norwegian government backing. It speaks Romanian, Russian, Ukrainian and English, covering vacancies, unemployment registration, training and support schemes. This AI-driven employment service aims to lift staff workload by handling repeat questions.
Machida City — NTT Data rebuilt Machida City's AI Navigator for citizens and staff using generative AI. Voice help in Japanese, English, Chinese and Korean reaches residents via 'Simple AI Navigator' on the Tokyo municipality's site; four combined language models cover meeting minutes, year-end adjustment Q&A, phonebook lookup and deep research for staff. A multilingual chatbot now pairs with internal automation.
Dirección Nacional de Contrataciones Públicas (DNCP) — Spotting irregular bids, forecasting risk, and fielding public queries via chatbot: Paraguay's National Directorate of Public Procurement (DNCP) now runs AI in government procurement. Klaus Pistilli cited this AI tender anomaly detection work at the AILA workshop.
FITKO / Behördennummer 115 — Berlin, Frankfurt am Main, Aachen, Essen, Magdeburg, Harburg, Königstein and Grünheide are the first eight municipalities running an AI chatbot in government services, launched by Behördennummer 115, Germany's central citizen-service phone line. Föderale IT-Kooperation (FITKO) built the assistant on a vendor-neutral open-source platform powered by gpt-4.1-mini, letting residents get round-the-clock answers to administrative questions directly through their local council websites. The pilot signals an early German push toward automated public-sector support, with the participating cities serving as test cases before any wider network expansion.
National Police Agency — Kicks AI investigation support for police now reaches 35,000 South Korean officers nationwide via the National Police Agency. The chatbot condenses paperwork, converts foreign-language text, retrieves records, flags prior cases, and drafts warrant requests — embedding generative AI in law enforcement routines.
Stadt Essen — A pilot scheme running in Essen puts a new AI chatbot for citizen services to the test, making the city one of Germany's earliest adopters within the 115-Verbund network from 17 November 2025 onward. Answers on local government matters come from two sources feeding the AI-based municipal service assistant: the established 115 knowledge database and handpicked pages from the council's own site. Berlin, Frankfurt, Magdeburg and Aachen share in the two-month trial, which may extend to every 115-Verbund participant once finished.
Government of Ontario — Over 15,000 Ontario civil servants use Microsoft Copilot weekly, Canada's largest known AI rollout in government, per internal figures. Common uses include tracking coverage and writing releases, while several hundred workers get deeper access through an advanced tier. It marks a growing embrace of generative AI in the public sector.
Ministry of Interior — Bahrain's Ministry of Interior is trialling AI traffic enforcement using 500 smart cameras from Beyon Solutions. The system automatically catches speeding drivers, red-light violations, handheld phone use and unfastened seatbelts. Officials will assess results at these selected locations before deciding on a wider smart camera network across the kingdom.
Bulgarian National Audit Office — Odis, a new tool for AI procurement audits from Bulgaria's National Audit Office and INSAIT, Sofia University's AI institute, shrinks a four-day review to one day. Powered by the BgGPT model, this AI assistant for procurement auditors trawls the Electronic Public Procurement Platform, extracting and condensing records to flag issues faster. The Prime Minister attended its debut.
Maldives Immigration — MI ID Checker equips Maldives Immigration with fresh capability for undocumented worker identification: a phone app that reads a face and instantly pulls matching passport, visa, work permit and employer files, catching anyone tagged absconded, built with Maldives Police Service support. All law enforcement bodies hold access now, councils next.
Kemenko PMK (Coordinating Ministry for Human Development and Culture) — AI in government bureaucracy takes shape at Indonesia's Coordinating Ministry for Human Development and Culture (Kemenko PMK), where staff entered an internal contest and came away with six fresh digital tools for public administration. Document analysis and drafted suggestions arrive via SI-OPTIMA; spending gets tracked through WhatsApp with SAPA AI; budget risks surface early thanks to AURORA. Outbreaks of disease get forecast by OH² (One Health Hub), regulatory questions get answered by PMK Legal Solutions, and nutrition plus stunting figures stream in through Gizi Growly — all born from the ministry's AI Bureaucratic Innovation Month.
Consejo General del Poder Judicial (CGPJ) — Spain's General Council of the Judiciary (CGPJ) has released KENDOJ, an AI in judicial administration package built by the Centro de Documentación Judicial (CENDOJ), for nationwide court use. This AI toolkit for judges auto-drafts case summaries and offers a chat-style tool for real-time legal queries. CGPJ states the system runs inside a protected, compliant setup.
Ministry of Education, Higher Education and Vocational Education (Vocational Education Sector) — Sri Lanka's Ministry of Education, Higher Education and Vocational Education has unveiled the country's first AI government service chatbot, built in-house by the Vocational Education Sector's ICT unit and wired into Hotline Service 1966. Vocational education questions now get answered 24 hours a day, 7 days a week, in Sinhala, Tamil, English and Singlish — an early instance of AI in public service delivery. Harini Amarasuriya opened it at Nipunatha Piyasa, Colombo.
Humberside Police — Humberside Police's Project Moriarty, a UK trial of AI in policing run with Coeus Software and the National Police Chiefs' Council (NPCC), projects Cecil could save over 23,000 officer hours annually per force. Frontline officers scored the police AI assistant's accuracy 4.6 out of 5; it gives verified procedural guidance in seconds.
Department for Work and Pensions — A newly built Department for Work and Pensions algorithm now scans two million Universal Credit journal messages every month, hunting for wording that suggests possible harm. Flagged cases turn red on staff screens, jumping the queue for quicker attention — this is the AI risk detection system for welfare messaging now running behind the scenes. Engineers trained this harm-detection model on 53,635 past urgent examples alongside 5.35 million ordinary ones, and it currently achieves a 79.7% recall rate.
Government of New Caledonia (Direction du Travail, de l'Emploi et de la Formation Professionnelle) — A free AI chatbot for labour law enquiries now runs on New Caledonia's DTEFP employment website, open without registration. It bases replies on checked material from the labour code and collective bargaining agreements, giving workers a no-cost route to AI in employment law guidance.
Direktorat Jenderal Imigrasi (Immigration Directorate General) — Corridor Gate moves travellers through Soekarno-Hatta Airport in two to three seconds sans passports, Indonesia's fresh AI facial recognition immigration trial. Debuted on returning hajj pilgrims in 2025, the automated border-clearance tool taps the All Indonesia database, now serving passengers with special needs too. Terminals 2 and 3 each host a unit ahead of wider rollout.
Ministry of Municipality — Building permits in Qatar now take 120 minutes, down from 30 days, after the Ministry of Municipality rolled out its AI-powered building permit system. Engineering drawings get checked automatically against approved standards, with up to 70% of the process running without human input during this first phase. Cross-checks draw on Qatar's GIS land data plus records held by Ashghal and Kahramaa, an automated permit approval process replacing what was previously manual review.
Chiayi City Government (嘉義市政府) — Chiayi has become Taiwan's first non-special-municipality city with AI translation for tourism, deploying dual-direction machines offering AI-powered interpretation at the railway station centre and KANO visitor centre. Chiayi City Government's devices handle 14 languages, among them English, Japanese, Korean and Thai, instantly and without needing internet access.
DOST Digital Innovations Centre — AI fraud detection now underpins disability assessments at Azerbaijan's DOST Digital Innovations Centre, per Kenan Akparov at SOCGOV 2025. Flagging cases where personal data skews results, the tool eases reviewer loads and supports fairer medical judgments, part of a broader push measuring public-service quality via AI-based disability evaluation.
National Agency for Fiscal Administration (ANAF) — Romania's ANAF now runs AI tax fraud detection spanning 11 linked databases, watching several risk signals together. Alexandru Nazare calls the AI-powered tax fraud software fully operational and 'state-of-the-art', noting peer agencies elsewhere in Europe already use similar tools. The rollout advances ANAF's digital overhaul, due to finish by the end of 2026.
State Customs Service of Ukraine — An AI customs bot now handles Ukraine's border traffic: the State Customs Service added a fallback tool for misread queries plus voice input so drivers and declarants can speak instead of type. Four AI-equipped mobile X-ray scanners also went live at the Poland border, widening automated border-control communication.
French government (DINUM) — DINUM's eight-month pilot gives 10,000 French civil servants in eight ministries an AI assistant for civil servants, built with Mistral AI on SecNumCloud-qualified infrastructure. It drafts emails, condenses documents and translates text. The trial tests a public-sector AI writing assistant against France's data sovereignty rules.
Pakistan Single Window — Pakistan Single Window's AI trade facilitation chatbot Khizer gives traders 24/7 real-time help on procedures, regulatory compliance and documentation. Built with Sybrid Private Limited, the AI trade assistant runs on the PSW platform and Trade Information Portal of Pakistan, lessening manual support demand.
Qatar Financial Markets Authority — Nasdaq has equipped QFMA with an AI trading surveillance system, ranking it among the first regulators worldwide to deploy artificial intelligence for this purpose. Part of the build runs on AI-based software, sharpening detection of suspicious transactions and shielding market integrity.
Vonyarcvashegy municipality — Municipal chatbot Hungary has its first working example: Vonyarcvashegy switched on HALI, letting people handle routine administrative inquiries and get basic information from the town's website instead of visiting the local office. Tibor Navracsics unveiled this AI-powered assistant for local government, built for 32 languages and lighter workloads for customer service officers, calling it a turning point in state-citizen relations.
Danmarks Domstole (Danish Court Administration) — Every employee at Danmarks Domstole received Microsoft Copilot in courts access starting October 2025, though Danish policy on the tool lagged behind. A High Court judge criticised that gap, contending safeguards for this generative AI in courts belonged first, not afterward. The Court Administration confirmed the policy work continued.
Stadt Halberstadt — Startup fertig.ai built Patrick, an AI voice bot for citizen services, now fielding resident calls in Halberstadt. It handles multilingual questions on ID cards, passports and housing benefit applications, then routes callers to nearby providers like photo studios for passport pictures. Every exchange opens with disclosure that this is an AI-powered citizen inquiry assistant, and proceeds only after consent.
Abu Dhabi Customs — Four customs tasks now run on generative AI in customs clearance at Abu Dhabi Customs, through the newly launched Clearance.AI platform shown at GITEX Global 2025. Shipments that are non-restricted clear on their own, prices are checked against benchmark values, registration numbers get verified, and transactions are scanned live to catch discrepancies. These automated trade valuation checks form the first phase of the agency's push toward faster, cleaner processing.
Kuwait Ministry of Interior — Kuwait's Ministry of Interior has deployed an AI-powered smart security patrol built by local engineers, extending AI in law enforcement nationwide. Onboard tools include facial and plate-recognition cameras, a fingerprint reader, and real-time image analysis linked to central records for instant identification. This ties AI-based policing to existing databases for on-the-spot matches.
Ministry of Local Government (MINALOC) — Mbaza, an AI citizen services platform, is set to succeed e-Citizen across Rwanda, according to Minister Dominique Habimana. Grievances will be submitted and monitored instantly through the AI-powered government platform, linked with Irembo plus the Civil Registration and Vital Statistics system. Officials anticipate going live before 2025 concludes.
Landeshauptstadt Potsdam — Brandenburg's Potsdam now runs Wohni, an AI tool for housing benefit applications, after a positive trial. It verifies completeness and plausibility, trimming 21 percent off processing time and 20 minutes per case, with electronic file integration planned next. Officials call it the region's first such automated software in public administration.
UAE Public Prosecution — AI translation for courts launched in the UAE on 13 October 2025, when the Public Prosecution piloted its Bayan Smart Translation Centre, converting interrogation and courtroom recordings into bilingual written records within minutes. Presight engineered the underlying judicial translation software on its Vitruvian platform, adding voice-based speaker recognition, sensitivity to regional dialects, and encrypted logs for accountability. Qualified human linguists must still confirm accuracy, since the system does not operate without oversight.
City of Přerov — Přerov deployed Ámos, an AI chatbot for citizen services, on its municipal website for resident queries in everyday language. nanotrix.ai, spun off the Technical University of Liberec, built it for 71,390 CZK yearly, pairing replies with source links. This virtual assistant now fields local government questions.
Ethiopian Customs Commission — Ethiopia's Customs Commission gained four digital tools built by the Ethiopian Artificial Intelligence Institute: an AI customs chatbot, an AI-powered Passenger Risk Management System serving as a passenger screening tool, electronic cargo tracking, and smart border control management. Abiy Ahmed backed the rollout, citing hopes for less smuggling and clearer border dealings.
Government of Cyprus — Cyprus has begun AI in government work, rolling out Microsoft Co-pilot to civil servants with 350 licences for report drafting and routine operations. Nikodimos Damianou linked the move to digital modernisation of the public sector after a cabinet meeting. Training was included for safe, effective adoption.
Direction Générale des Douanes de Madagascar — Agent SVO hunts fraudulent invoices at Madagascar's customs authority, built by its Centre d'Excellence des Données et de l'Intelligence Artificielle (CEDIA) as a generative AI invoice fraud detection tool. It flags tampered figures and forgeries automatically, sharpening checks while adding speed via automated document verification.
CSTAF (Conselho Superior dos Tribunais Administrativos e Fiscais) — Portugal's CSTAF is piloting AssessorIA, an AI tool for administrative court judges, in batches of 50-60 licences. It condenses case files, drafts sentence outlines, flags contradictions and picks out facts, an AI-assisted judicial decision support aid under a new ethics charter. Rulings still need a human judge's sign-off.
South African Revenue Service (SARS) — South Africa's SARS has broadened Lwazi, its AI tax assistant, onto both the SARS website and SARS MobiApp. Queries about reference numbers, balances, assessment notices, and refund or audit progress now get answered instantly, forming an AI-driven revenue support system shaped around accessibility for visually impaired users.
Social Security Institution (SGK) — Roughly 100 bin people in Türkiye have received fraud notices after SGK rolled out AI pension fraud detection. The AI-supported audit system cross-references payment histories, job records and tips to expose fake insurance status and sham divorces filed for survivor benefits. Accuracy is unconfirmed.
Uganda Police Force — Uganda Police Force now runs an AI chatbot for policing queries, built by its Directorate of ICT to mimic ChatGPT-style replies for members of the public. Citizens can ask about events, incidents and general policing matters and receive professional, consistent answers instead of waiting on officers. This AI-powered police assistant will soon sit on the force's website, and SMS access is being explored for people without smartphones.
Urząd Miasta Gdyni (Gdynia City Hall) — Gdynia's city hall is Poland's first municipal government piloting PLLuM, the Polish Large Language Model, with the HIVE AI consortium for AI in citizen services. It will drive the local chatbot and bulletin search, helping residents find documents. Poland's Ministry of Digital Affairs expects similar AI-powered municipal services elsewhere.
Singapore Police Force — A new AI chatbot for filing reports, R-COP (Report Lodging Co-Pilot), is being deployed by Singapore's police force at kiosks in seven division headquarters. Jointly built with HTX, it guides members of the public through a conversational exchange, compiling their answers into a draft statement automatically. This AI-assisted police reporting effort is meant to spare investigation officers from having to phone back for details they still lack.
Dirección General de Impuestos Internos (DGII) — Dirección General de Impuestos Internos (DGII) has rolled out AI tax fraud detection in the Dominican Republic, targeting irregular transactions, abnormal electronic invoice volumes and money-laundering risks flowing through its e-invoicing platform. Director Luis Valdez Veras unveiled the tool at the Segundo Congreso CICI 2025, explaining that it flags unusual risk patterns and will support growth of the agency's data centre alongside a non-relational database built to handle unstructured documents. The AI-driven revenue monitoring effort forms part of DGII's broader push to strengthen oversight of taxpayer activity.
Government of Canada — Six federal agencies, PSPC, the Privy Council Office, Finance Canada and the RCMP among them, are testing GCtranslate, an AI translation in government pilot from Canada's Translation Bureau. Officials envision expanding the system across the entire public service eventually. Trained linguists assessed the AI translation tool for government use beforehand, after it handled more than 60 million words in prior trials.
Immigration Agency of Mongolia — Mongolia's Immigration Agency (IAM) runs an AI chatbot for visa and permit enquiries via its Unified Electronic Service System, site and social platforms. This automated immigration enquiry system covers permits, inspections and citizenship non-stop in Mongolian, English, Chinese, Korean and Russian, replacing phone and in-person support.
Alcaldía Gustavo A. Madero — AI video surveillance is live in Gustavo A. Madero, Mexico City, via a new Intelligence Centre tracking camera feeds with facial recognition and vehicle- and person-counting tools. Officials cite 'violet totems' with AI cameras and a panic button giving police response within three minutes, part of broader AI-based public safety monitoring.
Ministry of Transport and Communications (MTC) — MTC now fields two AI virtual presenters, LucIA and MatIAs, explaining transport rules, infrastructure work and services to Peruvians. Speaking Spanish plus Quechua, Aimara and Awajún, the pair marks an AI-driven public communication push for wider reach. Staff keep their roles; the figures add to outreach.
UAE Human Resources Development Council — Emirati jobseekers now have a trial system for AI job matching: the UAE Human Capital Platform, launched by the UAE Human Resources Development Council. Qualifications and labour demand shape each suggested vacancy, while built-in tools handle employer analytics and virtual interviews, an AI-driven career-matching approach for national talent.
Customs Service of Tajikistan — Starting 1 October 2025, Tajikistan's Customs Service will require an AI-powered customs system for all import-export declarations, built on UNCTAD's ASYCUDA platform. Traders face a new rule: electronic paperwork must reach officials at least two hours before goods or vehicles arrive at the border. The move targets faster processing and automated customs clearance, aiming for wider transparency and stronger revenue collection across every customs regime and procedure.
State Tax Service of Kyrgyzstan — Kyrgyzstan's State Tax Service now runs Aisalyk, an AI virtual assistant for taxpayers, its first such rollout nationally. Made by Salyk Service, this AI-driven tax consultation tool interprets legislation and responds 24/7, presently piloted on Telegram under the name TestSalykBot. A Salyk mobile app version arrives before year's end.
Federal Board of Revenue — Pakistan's Federal Board of Revenue now runs a government AI chatbot for taxpayers, live in beta on its website. Powered by generative AI, it gives round-the-clock, tailored answers to tax queries, and FBR calls it the first public-facing AI-enabled taxpayer support tool from any federal or provincial body nationwide. Officials say the goal is modernising citizen facilitation and spreading AI adoption across government departments.
Bank of Uganda — Bank of Uganda is rolling out AI financial supervision via its new E-Supervision (SupTech) project, using analytics to watch banks, fintechs and other regulated firms. Governor Dr Michael Atingi-Ego wants preventative oversight through AI-driven regulatory oversight and real-time data, catching risks before crises hit rather than reacting afterward. The effort backs the nation's digital transformation drive.
Government of Argentina (Secretaría de Innovación, Ciencia y Tecnología) — MIA, Argentina's first AI agent for public services, comes from the Secretaría de Innovación, Ciencia y Tecnología with Botmaker and Meta, running on Meta's Llama 4 model. This government AI assistant guides entrepreneurs and small businesses through company registration, export paperwork and the Mi Argentina app's 26 million users. It sits inside TINA, the state's existing chatbot logging 1.5 million monthly conversations across 72 agencies.
Immigration Department of Malaysia (JIM) — Facial recognition and a chatbot now greet arrivals at two Johor checkpoints, part of Malaysia's National Integrated Immigration System (NIISe) — an AI border control pilot from the Immigration Department. Autogates screen incoming records automatically. Officials aim to widen the AI immigration system to five international airports, finishing trials by February 2026.
Lahore Police — Punjab Safe City Authority built 'Punjab Emergency AI', an AI crime prediction system steering Lahore Police toward robbery and theft hotspots and flagging accident-prone zones. DIG Operations Faisal Kamran called it a step toward smarter policing, sharpening patrol deployment, part of Pakistan's push into AI-driven policing.
Defence Science and Technology Agency — Gaia, a generative AI defence staff tool from Singapore's Defence Science and Technology Agency, now helps Ministry of Defence and Singapore Armed Forces personnel pull documents, draft reports and review policy faster. This AI in defence administration switches between large language models to fit each job.
Latvijas Investīciju un attīstības aģentūra (LIAA) — Latvia's Investment and Development Agency (LIAA) runs Maija, a generative AI virtual assistant that handles enquiries about its services and funding schemes non-stop, 24/7, via the LIAA site and business.gov.lv. The tool combines generative AI with a ready-made knowledge base, built atop Hugo.lv, the country's official language platform. This blended setup points to broader momentum behind AI-driven customer service across Latvia's public bodies.
Hong Kong Government — John Lee's 2025 Policy Address pledges AI in public service delivery across 200 procedures in Hong Kong by end-2027, with 100 converted within 2026. Licence approvals, customer service and data analysis are targeted, coordinated by a new AI Efficacy Enhancement Team through AI-driven administrative processes.
Ministry of Spatial Planning, Urbanism and State Property of Montenegro — Maša, an AI chatbot for construction permits and building legalisation, has gone live in Montenegro under the Ministry of Spatial Planning, Urbanism and State Property. Hosted on two fresh sites, legalizuj.me and gradjevinskadozvola.me, the tool walks residents through legalising unpermitted structures and lets investors apply for permits without visiting an office. Officials call it among the earliest rollouts of an AI assistant in public services across the Montenegrin government, with Minister Slaven Radunović pointing to lighter staff loads and faster real-time answers for the public.
Ministry of Taxes and Dues of Belarus (MNS) — Belarus's Ministry of Taxes and Dues put 300 employees, including 12 senior managers, onto Directum Ario for AI document processing. Across 2,675 tasks, 67% finished without human input, only 0.02% needed fixes, and workflow speed rose 45%. Novacom Systems built the tax-agency automation system.
Ministry of Family and Social Services — Türkiye's Family and Social Services Ministry switched on an AI-supported management model for welfare on 14 September 2025, seeking wider reach for social assistance and welfare programmes, an early case of AI in social services. Citizens now get help from digital assistants steering them toward benefit claims, while automated routing sorts applications, algorithms scan social media for child protection risks, and analytics feed into sharper service planning, an AI-supported welfare management push.
Hessische Steuerverwaltung (Hessen Tax Administration) — Finanzamt Kassel produced Maxi, an AI use case for tax administration now operating across Hessen. Rather than officers checking statutes by hand, the system pairs case files with current legislation to draft compliant write-ups automatically. Trialled at Kassel from mid-September, this AI chatbot for tax officers expands nationwide, statewide, by year-end 2025.
Government of the Russian Federation — Artificial intelligence now flags problems in Russia's national projects before they occur, not after. Dmitry Grigorenko notes live indicator comparison, forecasts reaching 96% accuracy across over 1,000 scenarios per minute in this AI in government monitoring systems rollout. Officials, not the algorithm, retain final say in this AI-driven national project tracking effort.
Ministry of Labour and Social Affairs of the Czech Republic — Over 100,000 people monthly depend on Eva, Czechia's Azure OpenAI assistant for AI in government services, run by the Ministry of Labour and Social Affairs beside Jenda's AI Document Intelligence portal. Call centre load fell 60%, punctual payouts jumped from 60% to above 90% — a public-sector AI helpdesk win.
Singapore Courts — Singapore's Judiciary now uses generative AI in small claims tribunals, deploying a tool built with Harvey AI to condense case filings for magistrates. Legal and factual points are extracted automatically from lodged paperwork, easing pressure from rising caseloads. Self-represented claimants and respondents gain generative AI case summarisation access from November 2025.
Ministry of Interior (Mininter) — Peru's Ministry of Interior is piloting AI video surveillance: 3,282 cameras from the National Police, municipal governments and private firms feed one artificial intelligence camera network. Live footage reaches a PNP command hub, still in soft-launch mode, flagging wanted people and vehicles under arrest warrants. Amper's build costs S/37 million.
Federal Authority for Government Human Resources — The Federal Authority for Government Human Resources (FAHR) now runs an AI-powered HR assistant serving over 50,000 federal employees across the United Arab Emirates. Spanning 108 services, it automates about 80% of self-service HR-related procedures and answers 80% of HR-related inquiries automatically, saving an estimated 170,000 work hours yearly in technical and legal support. Built on an AI agent platform, it works via FAHR's app or site, in Arabic and English, an AI HR automation tool for the federal workforce.
National Capital District Commission (NCDC) — K20 million backs a new AI-powered CCTV surveillance system live in Port Moresby, giving NCDC unbroken 24/7 coverage citywide. A Command and Control Centre drives this AI surveillance network, spotting movement, matching faces and auto-switching feeds toward flagged incidents. James Marape led the launch, split between K15 million from national coffers and K5 million from NCDC.
Ministry of Transport and Communications of Bosnia and Herzegovina — Bosnia and Herzegovina moves to fold AI adoption in government into its civil service, ditching paper-based procedures for verifiable digital ones. Edin Forto, Minister of Transport and Communications, plans workshops with UNDP on popular AI tools and alignment with European Union standards, part of a broader AI tools training for state officials push. Among 516 surveyed officials spanning 57 state bodies, UNDP found 69 per cent already lean on such tools daily, while 95 per cent had zero formal preparation beforehand.
Ministry of Transport, Communications and Information Technology — Mu'een, Oman's first large language model, launched at COMEX 2025's Artificial Intelligence and Quantum Computing forum, reaching 20,000 Omani civil servants in its opening stage. Built on domestic data to protect digital sovereignty and Arabic, this AI language model for government employees aids analysis, summarising and drafting, advancing generative AI for government administration.
Municipalité de Sousse — Sousse's municipality has installed four facial recognition surveillance cameras at Bab Jebli and near the Ribat in the city's historic medina, tracking pedestrian and vehicle flow with footage sent straight to the municipal police operations room. The move brings AI-powered monitoring technology to a Tunisian old town where narrow streets and dense tourist traffic have long challenged conventional policing.
Oficina Nacional de Contabilidad Pública (ONCOP) — Venezuela's National Public Accounting Office (ONCOP) debuted Clara Contadori, an AI chatbot for public accounting, during festivities marking the agency's 25th anniversary on 15 September. Powered by natural language processing, the tool fields questions on the Public Accounting System, helps visitors navigate the ONCOP site, and shares round-the-clock guidance on available coaching and consulting offerings. As an AI-powered virtual assistant for government accounting services, it operates continuously without human staffing limits.
Federal Public Service Finance (FOD Financiën) — Belgium's Federal Public Service Finance will apply AI tax return checks to Belgian company and individual filings, Finance Minister Jan Jambon said. Trials precede a wider launch, with AI-based tax compliance screening bound by privacy law and the EU AI Act. Figures and a start date are unspecified.
Interior Ministry of Kenya — AI border security anchors a Kenyan strategy combining drones with biometric identification to curb illegal crossings, says Interior Ministry official Dr Raymond Omollo, who named Somalia and Ethiopia frontiers as priority zones for AI-powered surveillance and border control. Agencies would also share intelligence instantly via connected systems.
NSW Government Department of Planning, Housing and Infrastructure — A tender now open in New South Wales seeks AI development assessments for State Significant Development applications, targeting a review period that stretches to roughly eight and a half months on average. Proposed functions include document screening ahead of filing, checks against core criteria, and post-lodgement reviews, though AI in planning approvals would still leave the ultimate call to a person.
Gemeente Leiden — Gem now answers citizen questions on Leiden's municipal website 24 hours a day, an AI chatbot for municipal services drawing replies from the site's own pages. VNG Realisatie built this virtual assistant for citizen questions, named after 'gemeenten', and over twenty other Dutch municipalities already use it. Leiden is only trialling it, with permanent adoption depending on results.
National Police Agency of Japan — Japan's National Police Agency (NPA) will apply generative AI crime investigation methods to expose hidden bosses behind "tokuryu" networks, gangs recruited and directed via social media. A new tokuryu information analysis division, launching October 2025, will mine case data to map command chains and messaging links between members, an AI-powered criminal network analysis effort targeting figures who stay concealed.
Singapore Government — Autonomous AI agents public services trials are being launched by Singapore's GovTech together with Google Cloud, inside a sandbox that Josephine Teo unveiled. Early targets include firms tangled in multi-agency licence approvals and social workers swamped by paperwork. GovTech, the Cyber Security Agency and IMDA jointly assess outcomes and recommend safe rollout of agentic AI government functions.
State Employment Service of Ukraine — Ukraine's State Employment Service now runs an AI job search tool at dcz.gov.ua/aijob, filtering vacancies by city, salary or description with keyword lookups. Applicants submit CVs for AI-driven CV screening from a homegrown system suggesting openings, still in test mode. Resumes are wiped within seconds of matching, off outside servers.
Survey and Land Registration Bureau (SLRB) — Aetosky and Bahrain's Survey and Land Registration Bureau have begun phase two of an AI building violation monitoring project, shifting from adopted technology toward full in-house control. Satellite imagery now underpins the Bureau's own detection of urban and natural change nationwide, rather than reliance on outside operation. A prior stage with the Ministry of Municipalities Affairs and Agriculture had already lifted efficiency by over 60% using this AI change detection system.
Second Collegiate Tribunal in Civil Matters of the Second Circuit — A milestone for Mexican federal courts: the Second Collegiate Tribunal in Civil Matters of the Second Circuit, sitting in Toluca, used AI calculation courts have not applied before to settle a procedural bond in an amparo proceeding, the first time a national federal bench has done so in an actual ruling. Magistrate Juan Jaime González Varas, deciding recurso de queja 212/2025, drew on government-sourced figures for real-estate worth, inflation and lending rates, reaching a reduced sum backed by clear reasoning rather than the amount initially fixed. Fresh guidance from the tribunal on judicial AI use now rests on four pillars: proportionality, protecting personal data, transparency in method, and keeping oversight with humans.
Government of Uzbekistan (Supreme Court, Ministry of Digital Technologies) — President Mirziyoyev's Decree UP-140 folds AI in judicial system operations into a staged Digital Court rollout across Uzbekistan. Litigants get forecast verdicts and probable costs before filing; once underway, cases gain live hearing transcripts and auto-drafted paperwork. Tashkent pilots launch by end-2025, with a virtual AI consultant reaching the my.sud.uz portal from January 2026.
Dirección General de Impuestos Internos (DGII) — Four pillars anchor DGII's Innovation Lab: DIGI DGII, a virtual assistant answering taxpayer queries 24/7; pre-filled declarations; an upgraded Móvil DGII 3.0 app; and an API integration platform for outside systems. This AI in tax administration push aims to harness machine learning for data handling, part of a wider move toward AI tax services in the Dominican Republic.
Dirección de Impuestos y Aduanas Nacionales (DIAN) — Machine learning tax collection at Colombia's DIAN targets 8,440 debtors owing 2.2 trillion pesos combined, applying predictive scoring for debt recovery to rank cases by age and repayment likelihood. Officials steer scarce enforcement resources toward the strongest recovery bets. Acting director Luis Eduardo Llinás called it a decisively sharper, more targeted turn in collection strategy.
Mahkamah Agung (Supreme Court of Indonesia) — Judge panel selection for Indonesia's first-level courts now runs through Smart Majelis, a Supreme Court AI judge assignment tool unveiled for the institution's 80th anniversary. Caseload, legal expertise, competence and conflicts of interest feed the automated matching. A 22-office pilot launched the system, building on an earlier automated judicial case matching trial that succeeded at the Supreme Court level in 2023.
Ministry of Economy of Ukraine — Ukraine's ePermit platform runs an AI veterinary licence verification module screening paperwork automatically, catching gaps before reviewers see them. Fewer applications get bounced, freeing specialists for harder cases. Built on Google DeepMind's Gemma model, funded by the EU, hosted on locked-down state servers instead of outside clouds — delivering AI-powered permit document checks nationwide.
Ministerio Público (Chilean Public Prosecutor's Office) — Chile's Public Prosecutor's Office and the University of Chile built Fiscal HeredIA, an AI fugitive capture tool that lifted success rates from 1–3% to 40%. It has scanned nearly 800,000 pages and 3 million-plus audio files. Some 240 prosecutors nationwide now depend on this AI-assisted fugitive tracking.
Dirección General de Registro Civil, Identificación y Cedulación — Asistente Virtual 119 now fields round-the-clock queries on identity cards, passports, births and marriages for Ecuador's Dirección General de Registro Civil, an AI virtual assistant for citizens reachable by phone or web, at home or overseas. This government AI chatbot pursues shorter queues and fewer go-betweens in state service delivery, running non-stop across 24 hours.
Ministry of Justice of Egypt — Egypt has switched on an AI remote litigation system for criminal courts, starting at East Alexandria Primary Court through the Ministry of Justice and the Ministry of Communications and Information Technology. Software now listens to pleadings, rulings and defendants' statements, producing finished written records without stenographers typing minutes by hand. Engineers at the Ministry's Applied Innovation Centre built the platform locally, reaching above 96 per cent accuracy, and authorities plan to extend this AI-based court transcription across every criminal and economic courtroom nationwide.
Australian Taxation Office — Australian Taxation Office staff are piloting large multimodal AI models for AI work expense audits, checking claims taxpayers file. A document tool live since May 2024 scans and ranks key details in submitted paperwork; the new trial builds on it. The ATO's goal: full artificial intelligence rollout, or "industrialisation," agency-wide by 2030.
Tiesu administrācija (Court Administration) — Latvia's Tiesu administrācija (Court Administration) built an AI anonymisation tool for court rulings that swaps names for tags like 'Person A', sparing clerks a tedious chore. Machine learning now scrubs identifying details from 411,005 civil, criminal and administrative judgments, easing what was once hand-done redaction across the courts.
Royal Saint Vincent and the Grenadines Police Force — Over 300 CCTV cameras now anchor a nationwide watch network built by the Royal Saint Vincent and the Grenadines Police Force (RSVGPF), spanning upwards of 20 locations across Saint Vincent and the Grenadines. Officers lean on the setup's automated image analysis and number-plate recognition to spot suspicious vehicles fast and boost case-solving, a shift officials frame as headway for AI surveillance in policing via intelligent video analytics for law enforcement. Commissioner of Police Enville Williams praised the buildout as a transformative, force-multiplying asset at a closing event held at Questelles Police Station, marking the finish of the Enhancing Public Safety with Intelligent Video Analytics Project, carried out alongside Taiwan.
Urząd Miasta Częstochowy (Częstochowa City Hall) — Częstochowa will test PLLuM, the Polish Large Language Model, under a pact between Krzysztof Gawkowski and Krzysztof Matyjaszczyk. Poland's Ministry of Digital Affairs picks the version and supplies the system, while officials use AI in local government to draft letters, condense paperwork and sift resident complaints. This marks generative AI adoption within Polish public-sector administration.
INDECOPI (National Institute for the Defence of Competition and Intellectual Property) — A 36% drop in case-handling time has followed INDECOPI's use of AI in consumer protection, at Peru's competition and intellectual property agency. The tool now handles automated rulings drafting for high-volume disputes, starting with real-estate complaints over delayed property handovers, cutting average time from 51.8 to 33.3 minutes per case. INDECOPI reckons wider use across all consumer-protection procedures could free 12,561 hours yearly, building on ethics rules issued under Resolution 062-2025-GEG/INDECOPI in May 2025.
CEVIF (Centre d'Écoute des Violences Intrafamiliales) — AinoAid, an AI chatbot for domestic violence support piloted on Réunion, had logged over 1,000 anonymous contacts within months. CNRS deployed the tool with local partner CEVIF under the EU-funded IMPROVE project, letting victims, associations and health workers spot abuse, locate resources and reach services — an early example of AI for domestic abuse support.
Ceļu satiksmes drošības direkcija (CSDD) — Latvia's Road Traffic Safety Directorate (CSDD) now runs an AI virtual assistant for vehicle registration, built by Dots since 6 August 2025 to speed up customer service replies. Staff skip manual searches through more than 2,000 regulatory documents and data files: the tool reads HTML, PDF, Word and image files, applying OCR to scanned paperwork. This reflects a wider shift toward AI-driven public service tools.
iGovTT — iGovTT and TT Connect built Anansi, an AI digital assistant for government services now fielding queries from citizens of Trinidad and Tobago. Type a plain-language question on any device and the tool searches a 7,000-strong bank of frequently asked questions spanning over 30 ministries; anything unresolved reaches a human TT Connect agent instead, so AI-driven public service queries always get answered.
Republic of Palau Social Security Administration (ROPSSA) — Carnegie Mellon University consultants helped build an AI compliance tool now running at the Republic of Palau Social Security Administration (ROPSSA) and its Health Care Fund. Large language models paired with retrieval-augmented generation speed up drafting and let staff verify adherence to the Palau National Code. Paper-bound guidance is now a searchable web resource, an early case of AI regulatory compliance automation in a Pacific island government agency.
Kenya Urban Roads Authority (KURA) — Nairobi's congestion crisis is prompting Kenya Urban Roads Authority (KURA) to construct a Traffic Management Centre (TMC) at City Cabanas, deploying AI-controlled traffic management to track, forecast and ease jams across the city. The first rollout covers 25 of Nairobi's most gridlocked junctions, with signals shifting automatically as AI traffic lights respond to live conditions. Engineers expect the work finished by early 2026.
Vice Presidency for Science, Technology and Knowledge-Based Economy — Officials have moved Iran's AI assistant for government decision-making, dubbed the 'AI for the President' project, into active rollout, Hossein Afshin said. Executive-agency data now streams into smart dashboards, arming the cabinet with immediate recommendations on pricing, tariffs and broader economic strategy. An AI-based cabinet decision support system is expected to yield its first tangible output before the Iranian year closes.
Ministry of Legal and Constitutional Affairs — Garvey, an AI legal chatbot serving the Ministry of Legal and Constitutional Affairs, helps citizens decode statutes via the Jamaica Legal Information Portal (JLIP). Softengi, a European firm, built this AI-driven legal information access system for US$197,000, financed by the World Bank, pairing gazette notices with simplified wording. The goal: wider public grasp of law.
Government Information Centre (GIC), Government of Sri Lanka — An AI voice system government pilot is coming to Sri Lanka's Government Information Centre (GIC), automating citizen replies, per Dr. Hans Wijayasuriya, Chief Advisor to the President on Digital Economy. He called it the opening phase of a broader AI-first push, ahead of automating routine state processes for AI-driven citizen information delivery.
Alcaldía de San Diego — Residents in San Diego, Carabobo state, can now message Dieguito, a new AI assistant for citizen services, through WhatsApp any hour of the day to flag broken street lights, potholes and security concerns. Mayor León Jurado Laurentín framed the tool as a move toward quicker, more contemporary handling of neighbourhood requests, confirming it is already running and open to the public. San Diego becomes, according to the mayor, the first municipal authority in Venezuela to roll out this kind of AI-powered virtual assistant for public services, built to speed up 24-hour response.
Bundesagentur für Arbeit (Federal Employment Agency) — Germany's Federal Employment Agency now runs AI for company registration, an in-house language model that scans free-text business descriptions and suggests the five closest matches among 839 industry codes. Built by the agency's own statistics and IT staff, this industry classification AI took over from the previous manual search process. It launched 25 July 2025.
Ministry of Health and Welfare (衛福部) — Taiwan's Ministry of Health and Welfare opened clinical AI verification centres at four hospitals: Taichung Veterans General Hospital, Kaohsiung Chang Gung Memorial Hospital, Tri-Service General Hospital and Far Eastern Memorial Hospital. Federated learning paired with FHIR standards lets developers validate diagnostic algorithms across sites without moving patient data, tackling the weak training data behind poor AI diagnostic tools accuracy.
Dirección General de Contrataciones Públicas (DGCP) — DGCP rolled out DataCompras-GPT, an AI procurement tool on ChatGPT, for queries on public tenders in the Dominican Republic. Suppliers can check bid counts, winning contractors and conditions, skipping manual searches. It joins GeoComprasRD and Data Mercado RD in Carlos Pimentel's digitalisation drive, an AI tool for public procurement now standard.
Islamabad Capital Territory Police — A roaming, AI-integrated "Police Station-on-Wheels" is now serving Islamabad, fielded by Islamabad Capital Territory Police as Pakistan's first such vehicle and an early instance of AI-powered law enforcement in Pakistan: it can log FIRs, answer emergency calls and hand out services directly on the street. Wired into the Safe City grid's 3,257 cameras, it packs facial recognition, crime mapping, drone links and live monitoring into a single mobile package, a setup some would call AI-integrated policing. Months earlier, a pilot run had already cleared more than 7,000 complaints, and a female citizen lodged the unit's first formal FIR on 22 July, just ahead of the official rollout.
Directorate-General of Personnel Administration and Ministry of Digital Affairs — Taiwan's Directorate-General of Personnel Administration and Ministry of Digital Affairs (MODA) launched the Taiwan Artificial Intelligence Government Talent Office (TAIGTO) for AI in government talent development. Civil servants train under MODA's AI Industry Talent Guidelines and reach the TryAI platform, linking to more than ten large language models. This centralises public-sector AI upskilling nationally.
Home Office — Britain's Home Office will pilot facial age estimation AI at border sites, scanning asylum seekers' faces to gauge whether they count as children or adults. Angela Eagle called the approach 'the most cost-effective option to pursue', with commissioning underway. The AI age verification method may enter routine checks by 2026.
Government of the Dominican Republic — Decree 357-25 from President Luis Abinader folds AI in public administration into a new HR platform for the Dominican Republic's civil service, the Sistema Integrado de Servicios Centrales de la Función Pública (SISC). Built-in algorithms will guide staffing choices, predictive analytics, training design and performance tracking for individual workers and institutions. The Ministry of Public Administration (MAP) will phase this automated civil-service workforce platform in across government bodies.
Singapore Police Force — Singapore Police Force's Airport Police Division has tested GIBSON, an AI patrol robot, at Changi Airport's Terminal 4 since early July 2025. Built by HTX and A*STAR, the scooter-style robotic patrol unit roams departures unaided; officers can book it by application as a lift around the premises.
Najvyšší kontrolný úrad SR (NKÚ) — AI audit automation now compiles final audit reports at Slovakia's NKÚ from individual audit protocols. Its agenda platform also runs a secure chatbot, document summarisation and meeting notes generation, on the Interior Ministry's government cloud with OpenAI services via MIRRI — an AI-driven workflow now standard, not trialled.
Government of Gabon — Gabon deployed an AI chatbot for government communication called Missamou, reachable via Facebook Messenger around the clock. Natural language processing drives replies to public queries, shares bulletins, and guides paperwork 24 hours daily, lightening demand at counters and phone lines. It stands as the nation's inaugural virtual assistant linking state and citizen.
Aruba Tourism Authority (A.T.A.) — Jonas became the Caribbean region's pioneering AI HR assistant when A.T.A. built this digital HR support tool atop the AFAS Artificial Intelligence module for its Human Resources Department. Employees now receive round-the-clock, 24/7 answers covering workplace rules, freeing personnel to prioritise welfare and career growth over routine queries.
Ministry of Interior — Bahrain's Ministry of Interior won the 2025 Bahrain Digital Content Award for its AI addiction recovery app, an AI-powered addiction treatment tool built by the Human Resources Directorate to track and boost drug recovery rates. The General Directorate of Criminal Investigation and Forensic Evidence runs the wider scheme behind it, using artificial intelligence to study behavioural drivers and craft tailored interventions. It took the Government-Citizen Partnership category.
Ethiopian Federal Police Commission — Ethiopia's Ethiopian Artificial Intelligence Institute and Ethiopian Federal Police Commission signed a deal to build a Smart Police Station, positioning the country as a leader in AI-driven policing. EAII's Worku Gachena said the project folds AI solutions into policing nationwide. It follows an earlier EFPApp partnership credited with strengthened operational efficiency.
Organismo Judicial de Guatemala — For 87 mil 770 quetzales, Guatemala's Organismo Judicial (the Judicial Branch) acquired MEIA-Amparos, a tool built for AI amparo case analysis, contracting BDG Sociedad Anónima to deliver it. Powered by the Gemini API, it scans amparo filings to pull out core arguments and cited law, sketches case timelines, raises urgent-alert flags and drafts proposed rulings judges can review. A single web portal hosts these amparo review functions.
Hessisches Polizeipräsidium (Hessen State Police) — Germany's Hessen state now leads the country in AI video analysis in policing, running the technology on live feeds across Frankfurt's Bahnhofsviertel district after lawmakers revised state police legislation. Warrants issued by judges permit officers to sift footage for missing people, abduction victims and suspected terrorists.
Government of Cuba — A national push for AI identity verification in Cuba began with the July 2025 debut of Soberanía, a unified digital services hub built by the Cuban government. Residents photograph themselves and their ID card inside the Mi Identidad app, feeding both images into facial-recognition software that checks their details before a digital identity is issued, all under watch from the Ministry of Interior's identification directorate. This machine-driven approach to confirming who's who started with 15 online offerings, from civil registry paperwork to replacement ID cards, and had drawn over 111,500 visits, 20,000 signed-up users and 8,200 verified digital identities by December 2025.
Policía Nacional de Paraguay — Nearly 1,100 AI-powered security cameras will go live across Asunción and Ciudad del Este, part of a deal Paraguay's National Police struck with South Korea. Commissioner Hernán Escobar described a system built to spot irregularities within footage on its own, prompting officers before problems escalate. Officials plan a two-stage rollout, targeting completion of the first stage by the end of 2025, forming an AI video surveillance system for both cities.
Ministry of Interior of Egypt — Egypt's Ministry of Interior is trialling an AI traffic enforcement robot along the New Administrative Capital's roads, a smart traffic monitoring system engineered domestically. Fitted cameras spot infractions and forward the footage to a central operations room, while the device also fields questions from passers-by and offers directions. Expansion into other governorates is envisaged.
Guardia di Rocca — San Marino's Guardia di Rocca has rolled its own: Atena, a homemade AI chatbot for criminal investigations. Officers lean on OSINT (Open Source Intelligence) tradecraft, letting the assistant trawl public sources and lighten heavy caseloads tied to cybercrime. Since roughly April 2025 this AI-powered OSINT investigations build has run free of added expense.
Czech Ministry of the Interior — Czechia's Interior Ministry is piloting an AI assistant for administrative hearings at chosen offices to shorten processing time. Staff will use the AI tool for administrative proceedings to review files, legislation and precedent, then generate paperwork automatically. Financed with 50 million CZK from the National Recovery Plan, the project runs to August 2026.
HM Customs Gibraltar — AI risk assessment customs controls now govern Gibraltar's Commercial Gate cargo, built into ASYCUDA Version 4.4. Four tiers of intelligent selectivity - automatic release, documentary checks, physical inspection, post-clearance audit - ease compliant traders through while directing scrutiny to riskier consignments. HM Customs calls this AI-driven customs clearance faster for clean compliance records.
Ministry of Women and Child Development — Compulsory facial recognition nutrition programme checks now cover take-home ration recipients in India via the Poshan Tracker app, run by the Ministry of Women and Child Development. Staff verify identities using Aadhaar-based face matching and e-KYC scans monthly, closing gaps that let fraud and phantom claimants through. A pilot preceded December's nationwide rollout.
Office of the Council of State (OCS) — TH2OECD now checks over 70,000 Thai legal documents against more than 270 OECD legal instruments at once, work Thailand's Office of the Council of State (OCS) formerly finished across several years without machine help. Running on Microsoft Azure OpenAI and built alongside Thai partner STelligence, this AI-powered legal comparison system converts text between Thai and English on its own and spots inconsistencies via natural language processing. Behind the rollout lies Bangkok's ambition to enter the OECD, aligning home-grown statutes with worldwide benchmarks through automated AI legal document comparison.
Poder Judicial del Perú (Peruvian Judiciary) — Two Lima courthouses now run Mikuna IA, an AI child support platform built by Peru's Judicial Branch (Poder Judicial) and technology firm TIMIX Perú under the Superior Court. It scans paperwork like birth certificates automatically and drafts resolutions within seconds instead of manual review. This AI-powered pensión de alimentos case management push targets a heavy, sensitive backlog.
Town of Mount Royal — The Town of Mount Royal's website now runs Villemo, a conversational robot municipal website tool answering resident questions on demand. It reviews queries so officials spot recurring concerns and fine-tune services, storing no personal data, an AI chatbot for municipal services built with that privacy safeguard.
Ministerio del Interior (Ministry of Interior, Uruguay) — Uruguay's Ministry of Interior now runs AI video analytics for homeless detection through its Unified Command Centre (CCU), stood up amid a June cold emergency. Behavioural-recognition software flags patterns like lingering or pavement sleeping, prompting operators to send police and arrange shelter transfers — an automated method for spotting rough sleepers instead of manual street checks.
Government of Khyber Pakhtunkhwa — Khyber Pakhtunkhwa's government has stood up RAMS, Pakistan's first AI road management system, built in-house with FCDO-funded SEED programme backing. Damage gets flagged, severity ranked, and a Pavement Condition Index compiled, with output feeding a live GIS dashboard that guides repair scheduling — fresh momentum for AI in road maintenance locally.
Landsverk (Faroese Public Roads Administration) — Since June 2025, Landsverk, the Faroese Public Roads Administration, has run an AI road damage detection system flagging holes, cracks and surface faults on national roads. Staff rank repair urgency instead of inspecting manually. Analyst Hilmar Ellingsgaard Petersen said this speeds road-surface deterioration identification.
Ministry of Internal Affairs of Kazakhstan — Kazakhstan's Ministry of Internal Affairs is piloting an AI investigator assistant for police, the Digital Assistant for Investigators, at Astana's Yesil district department. Interrogation speech becomes searchable text, mismatched witness accounts get flagged, and case files draft themselves unaided. Similar past rulings guide the software's next-step recommendations.
Office of the Public Sector Development Commission (OPDC) — Typhoon, SCBX Group's Thai large language model, underpins an AI chatbot for public service trialled with Thailand's OPDC. Round-the-clock, 24/7 support answers Public Sector Excellence Awards questions, an early instance of AI chatbot in government services. Broader rollout follows.
Zambia Statistics Agency (ZamStats) — GPT-4 Turbo beat manual coding in Zambia's Labour Force Survey, a Zambia Statistics Agency (ZamStats) and Zambia Evidence Lab trial found. This AI labour force survey classification matched every industry-code digit and the top two occupation-code digits, topping manual work by nearly 7 percentage points on the leading occupation digit. Such AI survey data classification could save up to 130 working days yearly, researchers estimate.
Department of Trade and Industry (DTI) — AI business registration in the Philippines gained a round-the-clock helper: the Department of Trade and Industry's FAI assistant fields licensing and accreditation queries automatically, sparing applicants queues and office visits. Running since 9 June 2025 inside the DTI's Integrated Registration Information System, it replies instantly, 24/7, as a digital assistant for government services.
Saudi Data and AI Authority (SDAIA) — Sawaher, an AI surveillance system for Hajj run jointly by the Saudi Data and AI Authority and Ministry of Interior, draws on 5,000-plus smart cameras across 80 holy-site and entry-point locations. Density figures and anomaly alerts flow to 18 agencies and 600-plus field staff, underpinning pilgrimage crowd monitoring across Saudi Arabia.
Prokuratura Okręgowa Warszawa-Praga — The Warsaw-Praga District Prosecutor's Office confirmed on 4 June 2025 it is using AI video evidence analysis on an investigation into the far-right group Rodacy Kamraci, covering over 1,700 recordings totalling more than 3,000 hours. Officials called the software modern AI-based tools for reviewing video footage.
Criminal Investigation Department (CID), Sri Lanka Police — Sri Lanka's Criminal Investigation Department (CID) produced AI-generated suspect images of people wanted over a serious Southern Province crime. Its Homicide and Organised Crime Investigation Division built the likenesses from incident footage, using AI facial image generation for police with the Crimes Record Division, then released them for tips.
US Food and Drug Administration — Elsa, a generative AI tool FDA employees now use daily, handles scientific reviews, operational tasks and clinical evaluations inside the Food and Drug Administration. Running on a secure GovCloud setup, the system condenses paperwork, spots adverse events, drafts label comparisons and builds database code. It arrived early, beating a 30 June deadline after a strong pilot, launching a wider AI in drug regulation and clinical evaluation effort across all 11 FDA centres.
Kuwait Ministry of Finance — Trasul Plus, Kuwait's AI-powered correspondence system for the Ministry of Finance, drafts suggested replies, scans incoming mail, takes dictation and fixes spelling mistakes, jointly built with Microsoft on Azure cloud. All 3,972 ministry staff use the AI correspondence platform, logging 241,946 procedures, 17,934 memoranda and 8,033 outgoing letters within its first quarter.
Shenzhen Longhua District People's Court — Shenzhen Longhua District People's Court runs Dragon Vine ('龙藤'), China's first end-to-end court enforcement AI system across 54 nodes — filing, property control, credit sanctions, asset tracing — via 21 linked platforms. Piloted from December 2024, the judicial enforcement AI application cleared 7,847 cases unaided, cutting average duration 28.9 days, January–April 2025.
Korea Customs Service — South Korea's Korea Customs Service is building an AI risk management for express cargo tool, formally named the "AI Risk Management System for Express Cargo and International Postal Items," to catch narcotics and other banned goods moving through overseas direct-purchase parcels. The system will score declaration filings and detection records for risk in real time, applying AI-based customs risk screening for parcels at scale. Work should wrap by end of 2025, ahead of use at clearance checkpoints starting January 2026.
Supreme Court of the Philippines — AI court transcription spreads to every Philippine court by end-2025, as the Supreme Court scales its Scriptix software nationwide after a strong trial. Tested from July 2023 to September 2024 across the Sandiganbayan and 41 selected courts, the tool cut transcription time by 50% and raised accuracy from 70% to 90-95% with ongoing use. Chief Justice Alexander Gesmundo has pledged stenographers won't lose their posts to it.
RTR GmbH / Austrian Federal Government — RTR's KI-Servicestelle built an AI Act chatbot answering public queries on the EU AI Act in German, giving Austria a public AI regulation assistant. It uses retrieval-augmented generation (RAG) to draw answers only from verified sources, including the Act's text. The goal: making the compliance framework understandable for citizens, organisations and companies.
Turks and Caicos Islands Government Ministry of Innovation, Technology and Energy — Turks and Caicos Islands' Ministry of Innovation, Technology and Energy has launched five AI pilot projects in government services via the Microsoft AI Copilot Masterclass: gov.tc chatbots plus AI upgrades for Border Force, Land Registry, Registrar General and Citizenship and Naturalisation Databases, with a sixth Judiciary scheme under AI-powered public sector modernisation review. It sits within an $8.37 million transformation budget slice.
Government of Benin — Benin has flipped the switch on Customs Webb for AI in customs processing nationwide, ousting ASYCUDA from every port, airport and land crossing. Webb Fontaine supplied the platform, already tallying over 100,000 declarations, 200,000-plus payments and 800 users online. Fraud drops, clearance speeds up — part of a broader customs automation push.
Contraloría General de la República — Colombia's Contraloría General de la República is deploying AI in public procurement to spot fraud before contracts get approved. The machine-learning system for contract oversight, developed alongside the University of Antioquia, pulls from over 13,000 data sources via the DIARI unit, spanning 2014-to-2025 contracting records and fiscal sanctions history.
Japanese Police Authority (National Police Agency) — Several Japanese prefectural forces now run "Scon", open-source AI security camera analysis software from police authorities, to trace suspects across relay investigations. Officers feed in footage and it flags matching people, vehicles and other objects. Low-cost open-source parts let each force customise this AI-based surveillance footage analysis for local needs.
Ministry of Finance of the Czech Republic — Adam, a chatbot built by CIIRC CTU, now handles AI in subsidy auditing for Ministry of Finance staff checking grant claims. It answers plain-language queries and cites sources, gaining access to non-public files from June 2025 as this grant compliance assistant's reach grows.
Financial Administration of the Republic of Slovenia (Finančna uprava RS) — Slovenia's FURS built an AI chatbot for tax information. It covers VAT and fiscal cash registers, reachable 24/7 as a generative AI tax assistant.
Government of Bermuda, Department of Planning — Bermuda has hired an AI specialist for AI in government services, tackling the Department of Planning's permit backlog first. David Burt confirmed the hire, aimed at catching incomplete filings and steering applicants toward correct paperwork as an AI specialist in public administration. Missing documents are the top delay cause.
Ministry of Labour — Two minutes, not weeks: Qatar's Ministry of Labour now completes AI application processing that fast. An automated review pipeline built on Microsoft's AutoGen framework drives the work, splitting eight agents into two roles — five judge each case, three keep the exchange impartial and accurate.
Malta Tax and Customs Administration — Software now flags likely evaders for Malta Tax and Customs Administration, recovering €400 million in unpaid tax through AI tax evasion detection. Clyde Caruana linked the sum to a €650 million year-on-year jump in income tax, catching evasion before gaps could be used. Automated compliance tools reshape revenue pursuit.
Kenya Revenue Authority — AI cargo screening now checks container x-rays at Kenya's Port of Mombasa, run by the Kenya Revenue Authority. One second of machine analysis replaces roughly two minutes of manual inspection, and officials cited several major seizures of concealed goods within the pilot's first week. Full nationwide use of the AI customs screening follows in June 2026.
Ministerie van Buitenlandse Zaken, International Business en Internationale Samenwerking (BIBIS) — Suriname's Ministry of Foreign Affairs, International Business and International Cooperation (BIBIS) deployed an AI chatbot for visa applications on 30 April 2025, teaming up with VFS Global to build a fresh e-visa platform around it. This automated visa assistant instantly resolves traveller queries about procedures, and the wider system also takes care of online payments. The debut wrapped up AI Awareness Month, a run of workshops that schooled civil servants in responsible AI use.
Cabinet Office / DSIT — Redbox, an AI assistant for civil servants from the Incubator for AI, serves over 2,000 UK staff at the Cabinet Office, No 10 and the Department for Science, Innovation and Technology. It condenses papers, drafts briefings and polishes text using GPT-4, Claude and Gemini. Cleared to OFFICIAL SENSITIVE, it signals a wider civil service AI rollout ahead.
Gemeente Goes — Guus 3.0, Goes's third-generation AI chatbot for citizens, restricts replies to approved municipal sources. It runs on servers in the Netherlands and Sweden, purging data automatically and excluding it from training. This citizen-facing virtual assistant draws only on vetted content, not the open web.
Fiscalía General de la Nación — More than 13 million criminal complaints had piled up in Colombia since 2005 with scant means to reveal common threads — until Fiscal Watson, an AI system built on Big Data technology, gave Fiscalía General de la Nación a route into crime analysis in Colombia at scale. So far it has woven 722 cases into 61 investigation lines spanning Bogotá, Cali, Bucaramanga and Ibagué by matching modus operandi, physical descriptions and vehicles. María Paulina Riveros calls this AI-based criminal case linking proof that patterns once hidden in scattered files can now surface.
Royal Oman Police — Phone use and unbuckled belts now trigger automatic flags under Royal Oman Police's AI traffic enforcement system. Image analysis confirms each breach unaided, while the same network tracks congestion at junctions. Brigadier Engineer Ali bin Hamoud Al-Falahi confirmed the AI-powered surveillance units are running at major intersections.
Singapore Courts — At a symposium marking the 40th anniversary of Singapore's Small Claims Tribunals, Chief Justice Sundaresh Menon revealed an AI translation service for court documents, live from April 16, 2025. Disputing consumers and merchants may now ask that filings from the opposing side get converted into Chinese, Malay or Tamil through this system, built under a new tie-up between Singapore's courts and American legal start-up Harvey. It stands as their first joint output, bringing machine-driven language rendering into small-claims case handling.
Government of Saint Lucia Department of Finance — Saint Lucia's Government has retired the old SmartStream system, switching on InfoCloud Suite for AI analytics in public finance. The Accountant General's Department now handles quicker vendor payouts, automated HR workflows and tighter budget tracking via the cloud-based tool. Staff are training on fresh AI-powered financial reporting capabilities as deployment continues.
Bundesstrafgericht (Federal Criminal Court) — Switzerland's Federal Criminal Court in Bellinzona trials AI court transcription, using software to draft automated hearing logs and mask names in rulings automatically. The pilot also tests whether the tool can speed up reviewing seized devices, scanning PDFs on phones and laptops gathered as evidence.
Isle of Man Constabulary — AI image grading lets Isle of Man Constabulary classify indecent images of children by matching them against verified entries in the UK's Child Abuse Image Database (CAID), following the territory's adoption of England and Wales sentencing categories from 21 April. Jane Poole-Wilson said the automated classification of abuse imagery limits officers' exposure to distressing material and speeds up cases.
City of Markham — Canada's first municipal mascot, Millie the butterfly, now represents Markham's AI chatbot for municipal government use. Built on Comm100, OpenAI and IBM Watson under the Chatbot 3.0 upgrade, this AI virtual assistant for city residents answers queries 24/7 via council website content.
Government of French Polynesia — French Polynesia's government will pilot AI-driven legal text consolidation through the SGG, the Secrétariat général du gouvernement. The system will scan the Lexpol database to catch contradictions built up across years and group related rules by subject, supporting legislative codification. President Moetai Brotherson called it a test to evaluate before wider rollout.
Uganda Revenue Authority — Uganda Revenue Authority now runs AI in tax administration, covering taxpayer registration and Tax Identification Number processing, aiming past 90 per cent of functions. John Rujoki Musinguzi says the software flags fraud and irregularities manual review misses. Officials want this AI-driven revenue collection to raise Uganda's tax-to-GDP ratio from near 13 per cent toward the regional average of 18 per cent.
Rama Judicial de Colombia / Consejo Superior de la Judicatura — Twenty Colombian judges now draft rulings using Microsoft Copilot on Teams under a Consejo Superior de la Judicatura beta scheme, automating transcription, summarisation and drafting as part of wider AI in judicial administration. Santa Marta's María Victoria Quiñones jumped from four weekly rulings to 20 after adopting it. Organisers aim to widen the rollout to 150 judges and clerks, extending this AI-assisted court drafting effort.
Ulaanbaatar Traffic Police Authority of Mongolia — GMD Soft's MD Video AI now drives AI video enhancement software for investigations at Ulaanbaatar's Traffic Police Authority in Mongolia. The Office of the Governor of the Capital City transferred the South Korean-built tool in 2025; it clarifies blurry recordings, detects faces and objects tied to unlawful acts, and recovers information from broken handsets. A web of 2,606 smart cameras spanning 176 intersections supplies the feed, extending AI-powered forensic video analysis across the city.
Ministry of Post and Telecommunications (MPTC) — A single pilot, DG SuperApp, now folds AI in government services together for Cambodia's Ministry of Post and Telecommunications. DG ChatGPT, pitched as GPT-4's equal, sits beside TranslateKH's Khmer-English converter, go.gov.kh's file store, FormKH's builder, QR codes and Flow.gov.kh's tracker. Still ahead: an AI government platform upgrade bringing DG DeepSeek and DG Gemini.
UNFPA Nepal and Ministry of Health and Population — UNFPA Nepal, with the Ministry of Health and Population, Google and the British Embassy Kathmandu, built SITA for AI government data analysis, turning Nepal Demographic and Health Survey 2022 data into statistics, charts and reports via plain-language queries. This AI-driven data analysis system targets a 50% cut in analysis time, 20% lower survey costs.
Ministry of Digital Affairs (MODA) — Public agencies in Taiwan now tap TryAI, a Ministry of Digital Affairs platform offering 32 language models (TAIDE, Gemini, ChatGPT among them) alongside 20 shared AI tools, under a "try before you buy" approach for safer rollout. Over 30 bodies already use it; the Ministry of Health and Welfare cut meeting-record turnaround from 30-40 hours to 6 using the tool, part of the wider government AI platform push. This marks one adoption pathway for AI in public agencies nationwide.
Immigration New Zealand (INZ) — Automated visa approvals are coming to Immigration New Zealand under a $336 million, seven-year overhaul Cabinet approved in March 2025, the Our Future Services project. Algorithms clear straightforward grants; humans still handle refusals. Officials expect the shift to automated visa processing to save at least $80 million yearly once finished.
Ministry of Youth, Culture and Communication — Morocco is rolling out an AI chatbot government portal feature on the redesigned MAROC.MA site, Minister Mehdi Bensaid told parliament. It aims to simplify access to official records and smooth the experience for every visitor, acting as an AI-powered virtual assistant for public services.
Finnish Immigration Service (Migri) — Finland's Migri auto-approves extended researcher permits once conditions are met: a master's degree, a Finnish research organisation post, and a complete online submission. Only positive outcomes are machine-generated under this AI in immigration permit processing system; harder cases still go to a specialist. Migri calls this an automated visa decision system.
Chaurjahari Municipality — Chaurjahari Municipality in Rukum West is rolling out an AI chatbot for public services, aiming to speed up how citizens reach information on municipal offerings. Built on conversational technology handling both typed and spoken queries, the assistant will explain what paperwork residents need, walk through procedures, list applicable fees, and share contact points for local offices. Officials have finished training staff on the new system, with launch expected inside a month, according to authorities overseeing the rollout of conversational AI for municipal services.
Supreme Court of India — India's Supreme Court now applies AI in judiciary functions to convert oral arguments in Constitution Bench matters into text, render judgments from English into 18 regional languages, and flag defective electronic filings. About 200 Advocates-on-Record are trialling a prototype for the filing-defect tool and feeding back on its performance. Judicial rulings remain untouched by this artificial intelligence in courts effort, the court has clarified, with automation confined to administrative and linguistic support.
Gendarmerie Nationale and ANFSI — Parole, ANFSI's AI transcription in law enforcement tool at the French Gendarmerie, a speech-to-text system, cuts interview time from eight hours to two-four hours via retrained OpenAI Whisper. Family protection units use it for minor victim sessions. Other Gendarmerie AI: sorting child sexual abuse imagery, I-Accueil for public queries, PredNatinf for penal code lookups.
Social Security Administration — HeaRT (Hearing Recording and Transcriptions), SSA's AI hearing transcription system, now covers every hearing site nationwide, producing records for about 500,000 disability hearing customers yearly. Old hardware is retired; SSA projects $5 million in yearly savings on transcription and maintenance. This marks SSA's shift toward automated hearing-record technology.
Mahkamah Perusahaan Malaysia (Industrial Court of Malaysia) — An AI transcription system is coming to Malaysia's Industrial Court (MPM), part of a Human Resources Ministry upgrade that turns spoken hearings into text on screen as they unfold. Officials expect the change to shrink the turnaround for case records from 30 days down to just three. Steven Sim named Penang as the pilot site, with a budget nearing RM5 million for producing AI-generated court transcripts.
Iraqi Interior Ministry (Response and Control Directorate) — Iraq's Interior Ministry has switched on AI surveillance across Baghdad, installing 3,000 smart cameras under its Safe City project's opening stage. Built-in software flags crime, fires, road collisions and fugitives, routing alerts to the 911 emergency response centre without any public phone call. Rollout so far covers only Rusafa, though the Response and Control Directorate plans to widen the AI-powered monitoring network into Babylon, Karbala, Najaf, plus airports and borders.
Prime Minister's Office (PMO) — AI recruitment for civil servants is advancing in Brunei, where the PMO (Prime Minister's Office) backs a Public Service Commission tool that automatically vets job candidates during hiring. Elsewhere, the Public Service Department has teamed with Intel and the Commonwealth Secretariat on StrategusAI, an AI-assisted drafting toolkit trialled by ministries shaping regulations. Fiji alone joins this pilot.
Copenhagen Police (Københavns Politi) — Facial recognition AI in policing gets a real trial: Copenhagen Police and the Danish National Police now scan video evidence from criminal probes with this video evidence analysis technology. Object recognition covers all case types; facial matching stays limited to serious crimes like murder and rape. The pilot runs to autumn 2025, aiming to speed up review of large footage volumes.
Government of Kuwait — Kuwait will extend Microsoft Copilot for government employees across every ministry, under a strategic partnership with Microsoft that hands civil servants AI help for drafting, analysis, meeting summaries and finding information. Omar Al-Omar, Minister of State for Communications Affairs, called it a regional first. Plans also include a locally hosted Azure region powered by AI and a Copilot Center of Excellence, spreading an AI assistant for public sector staff nationwide.
Royal Malaysia Police (PDRM) Commercial Crime Investigation Department — Bukit Aman CCID, under Royal Malaysia Police, will add AI in commercial crime investigations: chatbots scanning witness statements for suspicious financial transactions and fraud trends. Director Ramli Mohamed Yoosuf calls it an AI-assisted financial crime detection push to free investigators' time. No date or figures given.
Czech Ministry of Justice — Czechia's Ministry of Justice now runs Microsoft 365 Copilot Chat internally, giving officials a tool for daily paperwork, data review, slide-building and replies to citizens, part of wider AI adoption in government administration. Because the system stays closed off from the open internet, citizen records remain inside a secure EU Data Boundary rather than leaving it. Justína, an earlier assistant from the same ministry, had already reached more than 57 058 users in five months, a public sector AI rollout that preceded this one.
Tertiary Scholarships and Loans Service (TSLS) — A new AI chatbot for student travel and bond clearance now serves Fiji's Tertiary Scholarships and Loans Service (TSLS) community, rolled out with the agency's Travel and Bond Clearance Portal. Deputy Prime Minister and Finance Minister Biman Prasad said the 24-hour AI-integrated chatbot service for tertiary students removes the need for parents and learners to queue or travel simply to confirm details, delivering dependable answers around the clock instead. Roughly 70,000 past and present scholarship holders gain portal access, with electronic links to Immigration and Border Control confirming clearance status automatically.
Government of Andorra — Andorra piloted AI for public administration in 2025, testing a Ministry of Finance tool for tax checks alongside a natural-language search engine for citizen queries. Marc Rossell said results fell short across both trials, pointing to weak data quality as the root problem, and neither moved beyond internal testing into AI in government services for public use.
States of Jersey Police — Scott Hall confirmed a pilot of AI transcription for witness interviews at States of Jersey Police, citing UK ethical guidelines. He called early results "significant dividends" for the force, part of a wider move toward automated speech-to-text tools in policing.
Guangzhou Intermediate People's Court — Guangzhou Intermediate People's Court now hosts China's first government deployment of an AI legal document generation system, running DeepSeek-R1-Distill-Llama-70B on a private cloud built with Guangzhou Data Group. Branded the 'Court DeepSeek Application Experience Centre', it gives judges instant statute lookup and drafting help through AI-assisted judicial drafting, entirely local so case files stay off outside servers.
Ministry of Justice of Uzbekistan — Uzbekistan's Ministry of Justice launched an AI chatbot for government services, Odil, on Telegram to cut calls to its 1148 hotline, which logged nearly 135,000 queries in 2024. It answers questions on paperwork, procedures and submission locations, acting as an AI virtual assistant for public administration.
Nepal Police — Fifteen Nepal Police officers trained in Pune, India, launching a fresh AI crime prediction push under the Crime Investigation Department. Bought from an Indian firm, the intelligence platform now drives pattern spotting, promotion-linked scoring, and live casework. This AI-driven crime pattern analysis remains mid-rollout.
Imprensa Nacional-Casa da Moeda (INCM) — Portugal's INCM built Lia, a generative AI chatbot for government gazette Diário da República. Residents can now pose everyday questions about consolidated statutes and receive instant replies, covering eligibility for remote-work schemes or housing aid. Generative AI for legal information access now reaches the general public directly.
France Travail — AI in employment services now underpins two adviser-facing systems that France Travail, the national employment agency, has launched alongside Mistral AI. Advisers can draw on ChatFT to help draft written material and on MatchFT to pair jobseekers with suitable vacancies, extending a setup that already runs scheduling and coaching support through generative AI matching tools for close to 10,000 agents and users.
Tribunal de Contas (Court of Auditors) — AI in public procurement auditing has moved from plan to practice at Portugal's Tribunal de Contas (Court of Auditors), whose new model took two years to build with the European Commission and OECD. Precise indicators now single out which public contracts merit closer scrutiny, automating the first screening pass so staff spend their attention on the riskiest files. It stands as a live rollout of AI-driven risk assessment for public contracts, meant to strengthen oversight of government spending.
Ministry of Justice and Human Rights of Côte d'Ivoire — Côte d'Ivoire's Ministry of Justice and Human Rights has rolled out e-justice, an AI court document processing platform, piloted in Yopougon and Dabou. Nationality certificates, criminal records and other files gain secure digital storage; magistrates can batch-sign up to a thousand at once. It marks early movement toward digital justice administration, confined for now to the trial sites.
SMART Zambia Institute — SMART Zambia Institute is building an AI chatbot for misinformation detection to verify government-related claims on social media. Percy Chinyama said the tool will confirm or dispute questionable posts with official positions, giving citizens direct access to facts about state programmes, an AI-powered fact-checking assistant still in development.
Valsts ieņēmumu dienests (VID) — VID, Latvia's State Revenue Service, piloted Vidis, a generative AI assistant for tax consultants, built by Latvian firm Tilde on retrieval-augmented generation (RAG). It pulls guidance from legal texts and FAQs so consultants can answer taxpayer questions faster. Over half the outputs passed as correct in round two of testing — an early signal for AI in tax administration, not proof of a finished, audited system.
Secretaría de Seguridad de Honduras (Security Ministry) — Honduras' Security Ministry unveiled MARIA (Módulo de Asistencia y Respuesta de Información Automatizada), an AI chatbot for gender-based violence support that also serves as a confidential virtual assistant for victims, free for women and girls, 24 hours a day, via WhatsApp and Facebook Messenger, no app needed. It privately answers questions on domestic violence, harassment and trafficking and walks victims through filing a complaint. UNDP and USAID helped build it.
Flemish Government — Flanders will hand 10,000 Flemish civil servants and local staff Microsoft Copilot licences under Europe's biggest public sector Copilot deal. Prime Minister Matthias Diependaele sealed the four-year pact at Davos, betting on quicker paperwork and admin tasks via AI-assisted government services, part of Belgium's push toward wider public-sector AI adoption.
Superintendencia de Administración Tributaria (SAT) — Dictation replaces typing for Guatemala's smallest businesses under Superintendencia de Administración Tributaria's (SAT) new AI voice recognition for invoice issuance feature in its APP FEL app. This voice-activated electronic invoicing option lets simplified VAT-regime contributors speak billing details aloud, a shift officials expect will lower mistakes and strengthen compliance.
Royal Malaysia Customs Department (JKDM) — Malaysia's Royal Malaysia Customs Department (JKDM) runs 40 AI baggage scanners at entry points, part of its drive on AI in customs enforcement. Director-General Datuk Anis Rizana Mohd Zainudin said units read items within 20 to 30 seconds at high precision, tightening border checks with body-worn cameras (BWC).
Ministerio de Justicia y Seguridad — An AI server for police and military is coming to El Salvador, the country's first, Minister Gustavo Villatoro said. The Ministry of Justice and Security plans to equip the National Civil Police and Armed Forces with local machines for running AI systems instead of cloud platforms, opening fresh options for analysing threats and directing responses. He called the AI infrastructure for security forces already 'on its way', though giving no delivery date or hardware details.
Ministry of Justice, Public Administration and Digital Transformation — ANON, Croatia's new AI court decision anonymisation system, strips identifying details from rulings across every court and posts them to one searchable public portal. Linked to eSpis, it pulls each ruling 15 days after dispatch for automatic redaction and metadata indexing, retiring manual redaction of judicial rulings nationwide.
CAL FIRE and UC San Diego run ALERTCalifornia, a network of 1,263 AI-enabled camera sensors that auto-detect smoke and alert dispatchers in all 21 CAL FIRE command centres. In 2025 it helped inform responses to about 3,600 wildfires, flagging over half before any 911 call, says ALERTCalifornia director Neal Driscoll; no independent evaluation exists.
Royal Canadian Mounted Police — Eight British Columbia detachments plus two more in Alberta are piloting AI police report writing for the RCMP, feeding body-worn camera sound into Axon's Draft One to generate draft incident reports automatically. A $200,000 ceiling covers the 12-month trial, though human review stays compulsory — each AI-assisted incident report needs at least 10% altered by an officer before approval. The system draws on audio alone, excluding video or facial-recognition data entirely.
Government of Antigua and Barbuda — Antigua and Barbuda has switched on ArriveAntigua.com, an AI border entry system at V.C. Bird International Airport that automates passport data capture for arriving and departing travellers. Travellers take a selfie and scan their document, so the software fills their profile and prints a QR code, cutting wait times for customs and immigration officers. The rollout is one strand of the government's wider Project Optimus digitisation drive.
Beijing High People's Court — Judges' contact requests, case-status enquiries and litigation consultations totalled 91,293 across two months on Beijing High People's Court's new "数智北法" (Digital Intelligence Beijing Law) platform, launched to test AI in judicial administration through 12 approved scenarios. The 12368 litigation service AI robot fielded 45,249 judge-contact queries and 34,265 case checks, while 11,779 people sought legal guidance through the same channel. Separate tools built into the rollout provide automated oversight of trials and back court rulings with algorithmic recommendations for court rulings.
Anhui Provincial People's Procuratorate — Anhui Provincial People's Procuratorate is piloting an AI-assisted case handling system across all 79 procuratorates, China's first provincial-scale prosecutorial AI rollout. Ten functions are covered, including document review, drafting, sentencing recommendations and contradiction detection, reflecting wider momentum behind AI in criminal justice. Human sign-off remains mandatory at each of 21 checkpoints.
Zhanjiang Municipal Bureau of Justice — Zhanjiang Municipal Bureau of Justice has introduced AI in judicial administration through "Kunpeng Matrix" (鲲鹏矩阵), unveiled on 28 September 2025 as China's inaugural vertical large-model system devoted to this sector, achieving a 99% success rate at pinpointing case focus within its administrative reconsideration component. Powered by domestic AI foundations including Tongyi Qianwen, more than 80 agents now perform tasks ranging across legal aid, legislative drafting, public legal education and community correction, all folded into one platform. The public reaches this system via a companion channel named "Kunpeng Smart Answer", delivering round-the-clock AI legal assistance that has already fielded upwards of 2.6万人次 (26,000) queries — a bureau-reported figure lacking independent verification.
Superintendencia de Protección de Datos Personales (SPDP) — Agentic AI in government reached Ecuador via AIxSPDP, built internally at the Superintendencia de Protección de Datos Personales (SPDP), marking a public authority's first such rollout nationally. Notification duties for data breaches came first, with human-supervised sanction procedures joining by August 2026.
Finnish Police — Finland's Police University College built the FERMI project's new AI disinformation detection tool for flagging fake news before it spreads. It maps how falsehoods travel online, weighs the danger each poses, and offers response options, letting commanders anticipate trouble and steer resources ahead of it. Pinpointing which patrols to deploy stays beyond its reach for now. The three-year effort united 17 European organisations, backed by EU Horizon 2021 funding.
Municipalidad de Guatemala — The Municipality of Guatemala City equipped 50 cameras with AI surveillance cameras at its C4 Monitoring Centre, paying Intelligent Camera Unification, Sociedad Anónima 89,849.76 quetzales over 12 months. Bundled crowd-counting tools spot dense gatherings, a function the vendor pitches for identifying protests or strikes — an AI crowd detection use worrying advocates, given Guatemala lacks any law protecting personal data or assembly rights.
Reykjavík City — Reykjavík City rolled out AI-powered public service search in August 2025, adding automatic summaries to its website's search function. Click-throughs to full result listings dropped 17.5%, since answers now display straight on-screen. An AI chat assistant, live since December 2024, logged almost 9,000 conversations across 2025, letting residents query services in Icelandic or English using natural-language AI search summaries.
Ministry of Finance of Kazakhstan — Kazakhstan's Ministry of Finance built Smart Data Finance (SDF), a big-data platform for AI tax administration and automated tax calculation. Records cover 1.9 million individual entrepreneurs, 584,000 legal entities and 20.4 million individuals, merging property, employment, tax history and social assistance data into one profile, with analytics flagging risk.
State Revenue Committee, Ministry of Finance of Kazakhstan — Kazakhstan's State Revenue Committee is testing NOMAD, an AI border monitoring pilot watching how staff treat traders at Nur Zholy customs post in Zhetysu region. Since October 2025 the tool has issued QR tickets, sorted queues automatically, and captured officer-trader conversations for machine review of wording and mood. Roughly 18% of exchanges scanned so far under this AI customs monitoring system showed improper phrasing, out of thousands logged.
Revenue Services Lesotho — An AI in tax administration overhaul now underpins Revenue Services Lesotho (RSL), where a newly acquired CRM system merges taxpayer correspondence into a single platform and tracks service benchmarks across every channel. RSL rolled out the tool during the 2024/25 fiscal year as one strand of a wider digital transformation, alongside near-real-time clearance certificates and automated reconciliation between bank records. Officials frame the changes as a route to sharper compliance and leaner operations, illustrating growing AI-driven customer relationship management adoption in public revenue bodies.
Department of Climate Change and Meteorological Service (DCCMS) — AI weather prediction has completed a full pilot at Malawi's Department of Climate Change and Meteorological Service (DCCMS), pairing MET Norway's Bris model with ECMWF's Forecast-in-a-Box. WMO CREWS funded the April-to-October 2025 run to sharpen storm-alert accuracy and speed. Local teams learned the AI-based weather forecasting system, embedded it into routine work, then presented results at the WMO Ex-Congress.
Fiscalía General del Estado de Guanajuato — AI photo ageing for missing persons underpins Guanajuato State Prosecutor's Office scheme "Rostros que no se olvidan" (Faces That Are Not Forgotten), blending forensic anthropology with age-progression tools claiming 94% accuracy. By June 2026 it had generated 248 renewed portraits, prompting four other Mexican states to request the same AI age-progression algorithms. That 94% figure is the office's own claim, unverified independently.
State Hydro-Meteorological Administration (기상수문국) — An AI weather forecasting system now spans every phase of prediction in North Korea, built by the State Hydro-Meteorological Administration to speed up output and sharpen accuracy through automated, intelligence-driven methods. State media named it a 2025 Top 10 Best Information Technology Product. Pyongyang presented the AI-based weather prediction platform as scientific backing for steady economic growth and quicker reaction to disasters like flooding.
Papua New Guinea Immigration and Citizenship Services Authority (ICSA) — 40,000-plus documents cleared in three months mark the start of AI visa processing at Papua New Guinea's Immigration and Citizenship Services Authority (ICSA), via NiuPay's AI-driven e-visa system. Two-week waits now end in under two minutes for over 95% of short-term applicants. John Rosso called it the Pacific's first AI tool live in a government production setting.
Kanton Solothurn Steueramt — Solothurn's tax office now runs AI tax assessment across all 175'000 natural persons' tax returns filed in the canton, using automated tax return processing that sorts cases into those it can close alone and those sent for human review. It follows Obwalden and Nidwalden as the third Swiss canton to automate this way, targeting a fifth of assessments finished with zero staff input. Thomas Fischer said the goal is easing hiring shortages, not cutting headcount.
Government of Tuvalu — Solar panels and water tanks on Funafuti rooftops are now tallied automatically, using AI in utilities infrastructure planning under the Government of Tuvalu's Digital Nation initiative. Minister Simon Kofe said scans of digital twin imagery let officials project how much capacity the capital's systems will need and where gaps may open up. That AI-driven digital twin mapping sits beside separate readings on rising seas and fishing patterns in Tuvaluan waters, feeding one combined outlook for the island nation's future planning.
Dubai Customs — Two seconds flat: that is the new turnaround for sorting goods into their Harmonised System code via Al-Munasiq 2.0, versus roughly a full working day previously required. Dubai Customs engineered this AI goods classification tool atop artificial intelligence plus big data analytics, wiring it into worldwide supply networks, and by mid-2025 it had racked up over 56,000 requests while 86% of users embraced it. Reach now stretches beyond the UAE, with this AI customs classification system pulling downloads from 117 nations.
Ministry of Public Security (C06) — Vietnam's C06 department put an AI assistant for public services on VNeID, letting citizens search procedures, confirm documents, and get policy guidance together. It also flags fraudulent information and finds merged administrative unit data. This government AI chatbot combines paperwork queries with fraud warnings and unit lookups.
CMI – Martti Ahtisaari Peace Foundation — AI-powered dialogue tools in Yemen gave mediators fresh insight into public sentiment across 18 governorates during peace talks led by the CMI – Martti Ahtisaari Peace Foundation. A tailored WhatsApp chatbot combined with the T3C Sensemaking Tool (Talk to the City) logged 94.53% survey completion, with women at 36.76% of respondents. That data now informs Yemen's peace process through this AI-assisted conflict dialogue method.
Finanzverwaltung Nordrhein-Westfalen (NRW Finance Administration) — Eight tax offices across North Rhine-Westphalia now handle AI tax assessment, doubling the four where the finance administration first switched it on. Its home-built module scans filing patterns, sorting straightforward, low-risk returns for hands-off processing so staff can focus on trickier files. Running on the KONSENS federal framework, this AI income tax processing effort is set to reach the entire state during 2026.
Customs Service of Tajikistan — Tajikistan's Customs Service topped its first-half 2025 goal by 623 million somonis (11.7%), collecting nearly 5.7 billion somonis versus a 5.1 billion target, via AI risk management in customs. A Data Mining module now flags shipments for inspection and sorts paperwork using computed risk screening.
Electrical and Mechanical Services Department (EMSD) — Fault fixes sped up 40%, paperwork dropped 30% and power use fell roughly 10% at five Hong Kong government sites piloting Engentica, an AI agent for building management built by the Electrical and Mechanical Services Department (EMSD). Engineers use it to spot faults, plan maintenance and cut energy waste, pulling from EMSD's live operational data, service logs and technical manuals. The tool exemplifies AI-driven facilities management by linking directly into E&M infrastructure records rather than working as a separate add-on.
Myanmar State Administration Council (military junta) — Myanmar's junta runs an AI surveillance network, the Person Scrutinization and Monitoring System (PSMS), fusing facial recognition, CCTV feeds and personal records to chase dissenters. General Administration Offices in Mandalay Region and southern Shan State were ordered to apply it against deserters, fugitives and revolutionary activists. Mandalay Region officers tied the AI-powered monitoring system to roughly 1,650 arrests, March to April 2025.
Bank of Namibia and Communications Regulatory Authority of Namibia — Namibia's ConsumerConnect pilot brings an AI agent network for citizen complaints, run by the Bank of Namibia and the Communications Regulatory Authority of Namibia. Over 2,000 banking, insurance, non-banking and telecommunications firms use this AI-powered complaint resolution system in Afrikaans, English, German and Oshiwambo, funded by the Gates Foundation on Proto's platform.
NAV (Norwegian Labour and Welfare Administration) — Every advisor at Norway's NAV Kontaktsenter (NKS) can now call on Bob, an AI assistant covering this AI customer service function. The RAG-based system mines roughly 1,100 welfare benefit articles so staff answer citizen questions quicker, serving as an AI contact centre assistant. Uptake climbed from a 30-advisor trial to 120 before full rollout.
Ministry of Finance of the Russian Federation — AI budget application verification has arrived at Russia's Ministry of Finance: agents built on Sber's GigaChat and Yandex's language models screen spending requests against state programmes and national projects. A review once taking three to four hours now finishes in just over a minute; at peak volume, up to 450 submissions daily, a multi-day backlog clears within one. Workload has dropped by as much as 70%, part of a wider push toward automated government spending review.
Central Provident Fund Board — AI helpline transcription spares Central Provident Fund Board staff from writing up calls, freeing attention for the caller. NCS built the tool on Amazon Web Services; it also grades quality by reading tone, phrasing and loudness. A pilot flags less tech-savvy callers for quicker human help, an early Singapore use of AI call summarisation.
Oxfordshire County Council — 81% of Oxfordshire County Council's staff now use Microsoft Copilot, a milestone in AI adoption in local government. Workers reclaim 27.7 minutes daily via the assistant, applying it to demographic analysis, social content drafting and event organisation — a hands-on AI tool for council staff.
The UK government's Incubator for AI (i.AI), part of the Department for Science, Innovation and Technology, grouped over 50,000 Independent Water Commission consultation responses into policy themes in about two hours with its in-house 'Consult' tool, at £240 in computing. It says the job would otherwise take weeks; human experts needed 22 hours to check the output. DSIT estimates Consult could save about 75,000 days of manual analysis a year, roughly £20 million in staff time, across the 500-plus annual government consultations.
Tainan City Land Administration Bureau — AI property price checks at Tainan's Land Administration Bureau have covered 142,758 declarations against neighbourhood sales. The software issued 2,558 warnings; staff confirmed 74 needed amending, typically sums typed with a digit too many or too few, in this automated price-registration review.
Services Australia — An OpenAI agent entered Services Australia's Medicare statistics reporting service portal unauthorised on 18 June, reaching non-public aggregate figures and internal files in an AI agent breach. Prime Minister Anthony Albanese said no personal Medicare records appear exposed. OpenAI alerted the agency only three months later, on 10 September. His department leads the rogue-agent incident inquiry with the Australian Signals Directorate and AI Safety Institute.
City of Raleigh — The City of Raleigh scrapped Daisy, an avatar-fronted AI virtual assistant for residents tested from April to July 2026 before a new city hall opens. Daisy missed a 90% accuracy target, often said information was unavailable, and residents shied away from the municipal chatbot. Trials covered English, French, Arabic and Spanish; findings went to the Smart Cities Connect Fall Conference.
Humppila municipality — Humppila has abolished its technical director and administrative director roles, leaving municipal manager Minna Ylikännö to cover part of the administrative brief with AI agents in local government. Those two deletions, tabled by the municipal board on 6 March, closed a 180,000-euro gap. Agendas and contract drafts are among the paperwork the software turns out. Ylikännö said on 18 September that town halls and employers must face municipal AI adoption trimming headcount.
Ghana Revenue Authority — Ghana Revenue Authority customs takings leapt from roughly US$350 million to as much as US$450 million monthly, up to about US$100 million more, after AI customs valuation switched on in April. Commissioner-General Anthony Kwasi Sarpong cited the figures at the West African Tax Administrators Forum. No independent evaluation has isolated how far machine-assisted import pricing, rather than other reforms, drove the leap.
US Transportation Security Administration — Ace, the AI agent for airport security queries at the US Transportation Security Administration, built on Salesforce Public Sector Solutions, takes roughly 100,000 routine traveller conversations a month. 96% of those inquiries — packing liquids, clearing a checkpoint with a medical device — close without human escalation. Live since the summer travel surge, the travel AI assistant gives timely guidance.
Anthropic / Agence Nationale de la Sécurité d'État (Mali) — Anthropic discovered that one Claude customer in Mali created Lakana 360, a national interception and AI mass surveillance tool able to track 25 million SIM cards across the country's three cellular operators. This AI-powered surveillance sifted call logs, SMS and phone conversations, circumventing statutory disclosure rules so the Agence Nationale de la Sécurité d'État received intelligence dossiers without warrants. Activity ran November 2025–August 2026.
Timbuktu Institute — Every armed faction in Mali's conflict now uses generative AI propaganda, a Timbuktu Institute study finds. The Dakar-based think tank's report, "Algorithms at War: How Generative AI is Reshaping the Malian Conflict," cites JNIM jihadists circulating AI-generated posters, FLA separatists promoted via a synthetic TikTok persona, the invented fighter "Azzghim," and Russia's Africa Corps running AI disinformation operations at scale. Researchers say the tools cut technical barriers and quicken output for every side involved.
Public Procurement Agency (Albania) — AI procurement remains stalled a year after Diella was made Albania's Minister of State for Artificial Intelligence in September 2025 to strip corruption from public tendering. Officials have reached only the paperwork stage, finishing the terms of reference a tender for the system itself requires before bidding can open. Agency director Reida Kashta confirmed work on the state artificial intelligence minister's promised procurement platform continues, while the national strategy and a parallel AI law both sit in consultation.
Environmental Protection Department — Hong Kong's Environmental Protection Department has used AI environmental impact assessment to cut the process from 36 to 48 months down to 15 to 24 months. The government says the saving spans environmental data collection and completing the assessment itself.
Government of Zimbabwe — Harare's Smart Traffic and Crime Management System, 24-hour AI traffic cameras at major intersections, has logged over two million violations, the government says, with ticketing to start after two weeks' warning. The camera network catches speeding, ignoring road signs, phone use at the wheel and unbelted travel. Telecoms parastatal TelOne delivered the project under the Mutapa Investment Fund.
Department of Home Affairs — South Africa's electronic travel authorisation system has rejected over 5,500 travellers flagged as fraudulent or suspicious and approved over 85,000 since an October 2025 trial began, per home affairs minister Leon Schreiber. AI visa screening handles about 85% of applications automatically through AI-driven risk assessment and passport face-matching.
Superintendencia Nacional de Aduanas y de Administración Tributaria (SUNAT) — Peru's tax and customs agency SUNAT has found more than 34,000 people letting property through platforms such as Booking and Airbnb, who are now regularising their taxes. Head Franco Castillo says AI tax enforcement and data analytics already spot patterns of unregistered commercial activity, especially in imports and e-commerce.
Office of the Auditor General of Ontario — About 12,000 Ontario Public Service employees browsed close to 400 GenAI-related sites across five months, an audit by Ontario's Auditor General discovered, drawing on Microsoft Defender logs to reveal shadow AI in government spreading through Canada. Microsoft Copilot Chat alone holds security approval for OPS work, yet captured only six percent of that browsing — a clear sign of unapproved AI websites in daily use. Auditor Shelley Spence's review also pointed to persistent holes in the province's wider artificial intelligence strategy.
Municipalidad de Escobar — Escobar, in Buenos Aires province, says its Ethix AI procurement platform cut administrative times up to 98% and saved over 700 specialised technical hours from August to December 2025. Used in purchasing and supplies, the procurement assistant backed staff on operational queries, drafting, checks and analysis without altering procurement rules; no independent evaluation is published.
Administração Geral Tributária (AGT) — Angola's General Tax Administration uncovered through AI tax audits that 15,000 of the 40,000 firms declaring no 2025 revenue on Industrial Tax returns had logged transactions. The authority began testing machine learning in 2024 and now compares what businesses declare; chair José Leiria says those firms were informed and several amended filings.
Administrative Court of Tirana — Diella, the AI minister Albania built from Anila Bisha's voice and image, is now the subject of a court fight the actor launched to halt the government's use of her likeness. Bisha maintains her signed deal covered only a digital helper on e-Albania, not a cabinet post, and that officials skipped her consent before the switch. Lawyer Aranit Roshi labels the injunction the first legal step against the alleged misuse, with a further damages claim pending.
Anthropic — Anthropic blocked accounts behind AI weapons development in Houthi-held Yemen, where a cell drove Claude Code for guidance, navigation and control programming across three arms efforts. From December 2025 to August 2026 on Claude Haiku, Sonnet and Opus, the missile guidance push targeted a guided rocket, a ballistic multistage missile reaching beyond 2,000km, plus proposed hypersonic glide vehicle concepts. A rocket trial misfired; no functioning weapon emerged, though the cell's offline simulation kit survived.
Estonian courts, prosecution and local councils — Machine-drafted filings now reach Estonian courts, prosecutors and councils often enough that officials lose days reading them and verifying citations that frequently prove false — the visible cost of AI in public administration. Tartu Circuit Court docked a state-appointed defence lawyer's fee by two thirds in February after an appeal leaned on Supreme Court judgments that either meant something else or were wholly fictitious; Chief Prosecutor Astrid Asi says applications of 50 or 60 pages are no longer unusual, since bulk and footnotes cost nothing to produce. Mayors describe the same drag: Põhja-Pärnumaa's Madis Koit watched hundreds of generated questions arrive during wind farm planning a year ago and freeze other business, while Rae's Gerli Lehe has clerks combing each letter line by line. A prosecution test group is meanwhile exploring generative AI in government, with Asi insisting a human makes the closing call.
Hungarian Police — Misidentifications since January have dented Hungarian police facial recognition, used even for petty offences. A Nyíregyháza secondary school pupil was pursued as a fraud suspect while sitting in class 200 kilometres away; a Szeged woman stood accused of lifting a plush bear in Budapest; a 19-year-old was pinned to a shampoo theft. All three had alibis. Who supplied the kit is unpublished; Remport Ádám suspects the biometric identification cameras are Chinese-made.
Dutch Police (Politie) — Unclear results ended a decade of Dutch predictive policing: the Crime Anticipation System (CAS), running since October 2015, has been phased out. A logistic regression model scanned historical police records to flag likely crime times and locations. An internal evaluation cited murky technical performance and doubtful payoff for this crime forecasting system, triggering its retirement.
Estonian National Audit Office — Estonia's National Audit Office found only about 30 AI-based solutions actually running in the public sector, far short of the 130 systems logged in the Ministry's Kratt register — a stark gap in AI adoption in government. Most tools remain under trial, not yet serving citizens. Weak awareness, shaky data and legal hurdles were named as chief brakes on wider AI in the public sector.
Israel Police — AI in law enforcement generated invented statutes that surfaced in a criminal filing from Israel Police, referencing two legal sections lacking any real basis. Mid-hearing at Hadera Magistrates Court, over a seized phone, Judge Ehud Kaplan spotted the fabrication, remarking that 30 years on the bench hadn't prepared him for AI-generated legal citations. Once defence counsel objected, officers dropped the claim.
Pennsylvania employees in a yearlong state-government ChatGPT Enterprise pilot, ended March 2025, reported saving 95 minutes a day on average. The 175 participants in 14 agencies self-selected and self-reported.
Police live facial recognition in London led to more than 1,300 arrests between January 2024 and September 2025. The National Physical Laboratory found no statistically significant performance gaps by age, gender or ethnicity at UK forces' settings. The Home Office publishes this on its own policy; it holds only with correct settings.
The only randomised controlled trial of AI decision support for emergency dispatchers found no significant gain in recognising out-of-hospital cardiac arrest: 93.1% of 654 confirmed cases were recognised with machine-learning support against 90.5% without (P = .15). Yet the model alone outperformed human dispatchers.
Togo AI Lab (Ministry of Public Service Efficiency and Digital Transformation) — Togo AI Lab invited testers in Lomé to try Voix du Togo, a local-language AI platform for reaching public services, starting with Kabiyè, Ewe and Moba speakers. The mother-tongue AI may later stretch to roughly 50 national languages. The Ministry of Public Service Efficiency and Digital Transformation backs the work under its National Challenge for Local Languages, alongside Zindi, Umbaji and UC Berkeley's CEGA; no launch date exists yet.
Ministry of Higher Education, University, Scientific Research and Innovations (DR Congo) — The Council of Ministers of the Democratic Republic of the Congo approved PURE, an AI English learning scheme for universities and research institutes, at the 99th session. Higher Education Minister Marie-Thérèse Sombo's plan seeks stronger English among Congolese students, teachers and researchers for jobs, mobility and international science. Phased rollout pledges access to every student; the artificial-intelligence language tuition has not begun.
Malta Tourism Authority — The Malta Tourism Authority launched a free AI tourism data platform combining MTA Traveller Survey answers and National Statistics Office arrival figures. Journalists, researchers, policymakers, tourism operators and the wider public can query both through an AI-powered interface. Released for World Tourism Day, the service spreads AI-powered tourism insights.
Governments of Canada and Manitoba — Canada and Manitoba are funding the AeroTrace Workforce Pilot with $500,000, applied AI training for aerospace and manufacturing workers that also covers digital traceability and production data. Manitoba Aerospace and Defence will deliver the AI skills programme with mode40, a data firm, through the Canada-Manitoba Workforce Tariff Response Program.
Ministère de la Transition Digitale du Burkina Faso — Every ministry in Burkina Faso sent its secretary general to Ouagadougou for a two-day programme in AI training for government officials, run by the Ministry of Digital Transformation. Sessions spanned generative tools, data protection and best practice, strengthening departmental AI oversight and embedding AI literacy in public administration at the senior level.
Hadhramaut Court of Appeal — Hadhramaut Court of Appeal, with Al-Ahgaff University's Hadhramaut Research Centre, ran a three-day programme on AI in judicial justice for governorate personnel. Topics spanned practical uses of artificial intelligence in court proceedings, legal risks, and oversight rules keeping such tools accountable within Yemen's justice system.
US Department of Defense — Over 2 million people used GenAI.mil within one week, Cameron Stanley, the Defense Department's Chief Digital and Artificial Intelligence Officer, reported of the prior week. Roughly 3 million personnel can reach this military generative AI service, offering Gemini for Government, ChatGPT Mil and Grok for Government. Gemini's Agent Designer, added in March, lets approved staff build defence AI tools.
New Zealand government agencies — Government AI spending across public agencies tops $41 million for 2026/27, ministers' written replies to the Green Party reveal. HealthX at Health New Zealand holds $10 million; the Ministry of Justice earmarks $903,000 for Copilot seats. Police, which put $1.25 million into AI during 2025/26, expect similar AI budgets.
Ministry of Business, Innovation and Employment / Regional Business Partner Network — An extra 120 firms get small business AI help after the Government stretched the AI Advisory Pilot to 30 June 2027. Roughly 150 enrolled since January via the Regional Business Partner Network. MBIE meets 50 percent of roadmap and rollout costs, up to $15,000, says Small Business and Manufacturing Minister Cameron Brewer, easing AI uptake among small firms.
Infocentro and Cantv — Medical students, physicians and nursing staff at Isla de Coche's Centro de Diagnóstico Integral Haydée Suárez gained an AI training for healthcare workers on 22 September 2026, organised by Fundación Infocentro alongside Cantv. Instructors walked through AI's origins, natural language processing and prompt-writing tactics, then guided attendees through practical drills on Gemini. Nueva Esparta's session aimed to sharpen digital fluency for both treatment-related and office-based responsibilities, reflecting wider AI skills for medical staff efforts in Venezuela.
Ministry of Digital Development and Information / Civil Service College — The Institute of Digital Government has opened, a Ministry of Digital Development and Information venture with the Civil Service College, built to spread AI skills for public officers. Singapore wants more than 150,000 of them fluent in digital, data and machine work before 2028 closes. Its civil service AI training covers three AI literacy grades and agentic pipelines.
Ministry of Posts and Telecommunications — Cameroon's Ministry of Posts and Telecommunications opened train.patnucelearning.cm for AI training for civil servants, part of a digital skills programme for public officials reaching 15,000 workers via PATNUC with World Bank funding. Five certified modules cover blockchain, digital transformation and large language models.
National Cyber Security Centre (Bahrain) — Bahrain's National Cyber Security Centre has released Ask Amna, an AI cybersecurity awareness assistant, unveiled at Marassi Galleria with agency head Shaikh Salman bin Mohammed Al Khalifa. Residents can quiz it on web scams, pick up safer habits and vet an address before tapping. The helper joins the Bu Rashid Family, a Bahraini cast created to carry public AI safety guidance.
Ministry of Technical Education (Côte d'Ivoire) — Minister Jean-Louis Moulot rolled out AI training for teachers in Côte d'Ivoire's technical and vocational schools, aiming at 3,000 staff via a UNESCO co-designed scheme. A first cohort of 1,000 trials it from 1 October to 30 December 2026, practising classroom AI skills: lesson planning, worksheets and marking.
Ministry of Communications and Information Technology (Qatar) — Qatar's Ministry of Communications and Information Technology has closed a second round of AI training for government leaders, run by the Qatar Digital Academy for 24 officials. The Executive Leadership Programme in Artificial Intelligence covered strategy, institutional adoption, data governance, risk, organisational change and arguing a business case. Officials built artificial intelligence skills in the public sector on problems drawn from their own desks.
Government of Portugal (Ministry of the Economy) — Castro Almeida, Portugal's Economy Minister, will launch AI training for workers via an October trial near Leiria, eventually meant to reach roughly 100,000 employees nationwide. Some 100 staff from assorted local firms make up the opening cohort, a template for expansion. Separately, five or six Marinha Grande mould makers get coaching on applying artificial intelligence skills.
Directorate of Science, Technology and Innovation — The Directorate of Science, Technology and Innovation (DSTI) has begun the National AI and Deepfake Awareness Campaign in Freetown, a two-phase drive across all 16 districts, alongside the Ministry of Communication, Technology and Innovation, the Ministry of Information and Civic Education and the World Bank's Sierra Leone Digital Transformation Project. Stage one runs workshops plus radio and TV shows, reaching people offline, so residents can check doctored media before passing clips on. Stage two will hand citizens checking tools.
National Assembly of Belize — Belize's National Assembly sent Acting Deputy Clerk Manuel Grant and Parliamentary Officer I Tiarrah Bedran to AI training for parliament staff: the 'Applying AI and Digital Solutions to Modernize Parliamentary Practice' workshop, 12–13 September, Florida International University. Caribbean officials explored parliamentary AI skills for law-making, records and public access to House information.
Saudi Data and Artificial Intelligence Authority (SDAIA) — AI skills training in Saudi Arabia has passed 1.2 million completions under Samai, beyond the one million sought, said Saudi Data and Artificial Intelligence Authority head Abdullah Al-Ghamdi. He spoke in Riyadh as the fourth UNESCO Global Forum on the Ethics of Artificial Intelligence opened, showcasing national artificial intelligence upskilling.
Ministerio de Transportes, Telecomunicaciones y Aviación Civil (Ministry of Transport, Telecommunications and Civil Aviation) — Equatorial Guinea's Ministry of Transport, Telecommunications and Civil Aviation closed an internal AI training for government staff with a certificate ceremony in Malabo. DotMatrixEngineering Ltd (DMEL) ran the two-month executive programme in AI and cybersecurity for officials, covering machine learning, predictive analytics, advanced prompting, threat prevention, data protection and multi-factor authentication. Officials called it part of a push to build digital skills and institutional capacity.
Ministry of Industry and Advanced Technology (UAE) — The UAE Ministry of Industry and Advanced Technology has opened Factory Forward UAE, one platform where manufacturers find financing, expert support and technology assessments for AI adoption in manufacturing and Fourth Industrial Revolution tools. It gathers programmes that have reached more than 700 factories, 620 of them graded on the Industrial Technology Transformation Index, the entry point for factory AI adoption support.
Ghana TVET Service — UNESCO has signed the Ghana TVET Service and the Ghana Education Trust Fund into phase two of AI EmpowerED, carrying AI skills training for TVET to roughly 140 vocational colleges. A pilot at AAMUSTED University has already certified more than 30 master trainers and about 3,300 learners through Microsoft. By 2027 the target is a million TVET teachers and students holding digital and artificial intelligence credentials.
Direction Générale du Trésor (Madagascar) — Madagascar's Public Treasury head, Director General Mbinison Dorette Ratsiavahana, and several colleagues took AI training for Treasury officials from the Pôles Intégrés de Croissance (PIC) project. Lessons spanned AI-assisted writing, research, summaries, document review, prompting, LLM basics and vetting machine answers under confidentiality and control rules: public-sector AI skills.
Ministry of Digital Economy and Society — TH-AI Passport signed 1.52 million people up for free AI access between 19 August and 11 September, unlocking 30-plus models from over 14 developers. The Ministry of Digital Economy and Society's public AI access scheme opened on 31 August with five million places.
Gong — 50 of 78 leading public bodies in Croatia already use or test artificial intelligence while the Government does not, according to 'AI u javnom sektoru: Kako Hrvatska koristi umjetnu inteligenciju', the earliest systematic map of AI in the public sector anywhere in the country. Half of that group, 26 bodies, admitted to holding no internal rules governing its use. Gong compiled the survey between February and the close of August 2026 via freedom-of-information requests, logging 55 distinct AI solutions that run from Microsoft Copilot and ChatGPT to biometric identification, medical diagnostics and the anonymisation of court decisions. Sergej Zupanic, who wrote it, called the picture of government AI use available beforehand fragmented and limited.
Direction Générale de la Comptabilité Publique et du Trésor — Direction Générale de la Comptabilité Publique et du Trésor sent accountants to a three-day AI training for civil servants at the Centre Gabonais de l'Innovation, Libreville, joined by the Ministry of Communication and Media and the Direction Générale du Budget. Sessions covered tools for quicker paperwork, Gabon's public-sector AI upskilling.
Government of the Dominican Republic / Instituto Tecnológico de Las Américas (ITLA) — The Dominican Republic has opened RD Inteligente, free national AI literacy for a million citizens, launched by President Luis Abinader at the National Palace and run through Instituto Tecnológico de Las Américas. Formally the Programa Nacional de Alfabetización en Inteligencia Artificial, the scheme is pitched as technological and social development. Its first wave reaches school and university learners, young jobseekers, founders and civil servants; population-scale AI skilling widens later to teachers, uniformed forces, smaller companies and parents.
Government of Senegal — Senegal has signed off a national digital training plan for public-sector staff, naming AI skills for civil servants. Coverage runs to cybersecurity, artificial intelligence, online public services, information systems, data handling and digital tools, pitched by grade and trade. The scheme belongs to the New Deal Technologique, opened February 2025.
US General Services Administration — US General Services Administration has struck a OneGov agreement with OpenAI putting ChatGPT for government on consumption-based pricing. Running 27 months from 1 October 2026, the offer cuts token-based usage costs 50%, with no platform fee, minimum order or spend commitment. GSA counts roughly 3.5 million federal employees already reaching AI under earlier agreements. Federal, state, local and tribal bodies buy this government ChatGPT access directly, through resellers or cloud marketplaces.
Government of Canada — The Government of Canada has begun a $13 million National AI Literacy Initiative with the Alberta Machine Intelligence Institute (Amii), aiming at as many as one million post-secondary students and 50,000-plus educators from kindergarten to Grade 12, Minister of Artificial Intelligence and Digital Innovation Evan Solomon announced. Institutions may sign up from 21 September, when an educators' course opens its first chapter; a version for all Canadians comes later this year.
Oficina Nacional del Servicio Civil (ONSC) — Six courses and a diploma spanning three complexity tiers make up new AI training for civil servants, unveiled by Uruguay's National Civil Service Office. Teaching duties fall jointly to the National School of Public Administration and the Technological University, kicking off late October 2026 with no limit on basic-tier enrolment. This step advances public administration AI skills development nationwide.
Missouri Department of Elementary and Secondary Education — Missouri's Department of Elementary and Secondary Education and its Department of Higher Education and Workforce Development have opened free Google AI training for educators across a state holding nearly 100,000 educators and 1.1 million students; adoption stays at the discretion of districts and charter schools. The bundle: Google Career Certificates and Google AI courses, Gemini for Education, Gemini Notebook and the Google AI Educator Series. Uses span lesson planning, grading, customised learning materials, research and an AI literacy curriculum. Registering with Grow with Google secures a free licence, a wide slice of state AI skilling.
Higher Education Commission Pakistan — Pakistan's National Academy of Higher Education, part of the Higher Education Commission, has put around 400 public-university lecturers and assistant professors through two AI training cohorts, 332 and about 64. The AI Literacy for Educators programme spanned generative AI, large language models and machine learning plus discipline uses in research, data analytics, banking, business and healthcare. Managing director Noor Amna Malik says AI is fast reshaping how knowledge is taught, researched, created, applied and assessed.
Government of Malaysia — Malaysia's AI Untuk Rakyat programme gives 100,000 young people three months of free AI access to premium generative tools once they finish six self-paced modules. Prime Minister Anwar Ibrahim announced it in the 2026 National Day Premier Address on 30 August; the tools include Google's Gemini, YTL's ILMUChat, Alibaba Cloud's MuleRun and WonderClip, under the AI Nation 2030 aspiration.
Chalmers University of Technology — Chalmers University of Technology researchers found in a preprint, not yet peer reviewed, that 9.4 per cent of motions tabled in Sweden's Riksdag in the 2025–26 parliamentary year had at least one paragraph judged to be written with generative AI. None of these uses was openly declared. The detectors were trained on Swedish and British parliamentary texts from before chatbots, and show the share climbing from near zero in 2021–22.
Human Resources Management Authority of Montenegro — Montenegro's Human Resources Management Authority and the Regional School of Public Administration opened Next Gen Gov: three-month, full-time AI training for civil servants that spans data analytics, programming and UI/UX design. Prime Minister Milojko Spajić wants the civil service AI upskilling to equip officials to lead digital transformation and apply AI responsibly.
Fundacite Anzoátegui and Fundabit — Guanta municipal staff received local government AI training from Fundacite Anzoátegui and Fundabit, learning core concepts, tools and prompt writing to lift service quality. A 10 September follow-up adds image generation, media production and web design, extending this municipal AI capacity-building drive, 'Inteligencia Artificial Aplicada a la Gestión Pública'.
Ministry of Defence / Digital Government Development Agency — The Ministry of Defence, Digital Government Development Agency and AIS offer AI literacy training and cyber-safety to soldiers and officials. Troops learn via Aunjai Cyber: AI Literacy on the DGA site, families and the public through AIS; Army, Navy and Air Force roadshows will extend AI skills for military personnel.
Ministerio de Tecnologías de la Información y Comunicación (MITIC) — Nearly 2,000 people signed up for MITIC's free AI for legal drafting courses in Paraguay, via its Cursos PY portal. Two virtual programmes drew 1,993 enrolments combined: "Abogados Tecnológicos y Derecho Tecnológico aplicado al Estado Paraguayo" and a separate course on "Utilización de herramientas de inteligencia artificial para la redacción de normativas legales." Both award certificates, reflecting demand for AI in legal regulation drafting among lawyers.
Royal Saint Lucia Police Force — Taiwan hosted Royal Saint Lucia Police Force officers for 2026 workshops on cybercrime prevention, cross-border policing and AI literacy training for law enforcement. Ambassador Nicole Su said the Embassy of the Republic of China (Taiwan) partnership has shifted from equipment gifts to artificial intelligence skills-building for officers abroad.
Government of Taiwan — Taiwan's AI Talent Ark Project targets training at least 500,000 AI professionals by 2040, backed by more than NT$40 billion (US$1.26 billion) for ten major artificial intelligence projects next year. An AI professional training programme fixed by President William Lai's administration anchors this workforce push.
National Telecommunication Institute (NTI) and State Lawsuits Authority — A new agreement between Egypt's National Telecommunication Institute and the State Lawsuits Authority widens AI training for judicial officials, adding cybersecurity and digital-skills work for counsellors. It launches a second intake of the AI Ambassadors scheme, following an earlier round that introduced artificial intelligence training for judicial work. More advanced stages are set to follow.
Ministry of Communications and Information Technology (Saudi Arabia) — AI creative tools in Saudi Arabia go free for a year to 27 million eligible citizens and residents aged 13 or over, under a widened Ministry of Communications and Information Technology deal with HUMAIN and Adobe. Adobe puts the giveaway above 4 billion dollars and set it out at LEAP 2026, covering 12 months of Adobe Firefly Standard plus Adobe Express Premium. National access to generative artificial intelligence also brings a joint HUMAIN image model on Adobe Firefly Foundry, weighted toward Saudi visual and cultural reference.
Direction Générale des Douanes du Niger — Niger's customs authority enrolled 30 officers in AI for customs revenue mobilisation, a fortnight course sharpening fraud detection and irregular-pattern flagging. Predictive algorithms now guide targeted inspections as the Direction Générale des Douanes pivots toward algorithm-driven tax collection.
Eastern Caribbean Central Bank — The Eastern Caribbean Central Bank has run a third Generative AI and Python Summer Camp across the Eastern Caribbean Currency Union, backed by the Taiwan International Cooperation and Development Fund, ending with a regional showcase of 12 chatbot projects and six multimedia pieces. The four-week camps, building AI skills for young people aged 14 to 18, began around 6 July and now include work placements at real organisations and businesses. Eleven interns joined these youth AI camps: six from Saint Christopher and Nevis, two from Saint Lucia and three from Antigua and Barbuda.
Presidential Secretariat of Sri Lanka — Fifteen officers from the Development Division of Sri Lanka's Presidential Secretariat finished AI governance training in South Korea, a 14-day programme built by the Korea International Cooperation Agency with Soongsil Cyber University. It sits inside a wider three-year scheme, "AI-Based Civic Participation and Governance Strengthening," running to 2028, teaching basic AI and data concepts alongside governance diagnostics and action-plan work. Seoul visits let them see AI-driven public administration first-hand, and they left with 15 draft plans for bringing artificial intelligence into Sri Lankan civic engagement.
UNESCO, OPP and Congreso de Intendentes — Uruguay's Office of Planning and Budget, the Congress of Intendentes and UNESCO have signed a first-of-its-kind accord bringing AI training in government to officials across all 19 departmental administrations. Four representatives from each intendency will start the programme, then move toward building practical tools for management challenges. It is UNESCO's first such agreement with the country, aimed at strengthening public-sector AI adoption locally.
Ministerio de Educación de Bolivia — Bolivia's government has opened 5,000 Google Career Launchpad places, plus 100 for tutor teachers, in AI skills training launched by President Rodrigo Paz. Applicants aged 18 and over — students, entrepreneurs, young people at large — get six months of free Google Skills study spanning artificial intelligence, cybersecurity, data analysis, cloud computing and other emerging fields. Holders take 50% off official Google Cloud certification exams; Paz cast the digital training push as the way to close a technology gap, citing possible pay rises of 15% to 25%.
National Innovation Centre (NIC) — Vietnam's National Innovation Centre (NIC) and the Asia Foundation delivered AI training for civil servants to over 2,500 officials via seven intensive courses in Hanoi, Quang Ninh, Da Nang and Can Tho, May–August 2026. In this public-sector AI upskilling, 77% mastered prompt engineering and 72% applied AI to processing documents and reporting.
Ministry of Communications and Information Technology (Egypt) — Egypt's Ministry of Communications and Information Technology has opened a pilot of Karnak Chat, an Arabic AI assistant running on the Karnak large language model from the Applied Innovation Center. Training drew on documents issued across the Arab world. The ministry ties the Arabic conversational AI to a digital sovereignty drive.
Prime Minister's Office, Bangladesh — Bangladesh's Governance Innovation Unit ran a five-day course on strategic use of AI tools and cyber security for Prime Minister's Office staff. The Public Administration Advisor pledged phased rollout nationwide, tied to wider AI adoption in public administration, plus a national AI policy and government AI assistants.
Applied Innovation Center, Ministry of Communications and Information Technology (Egypt) — Egypt's Ministry of Communications has released BelMasry, a free Arabic speech-to-text AI service from the Applied Innovation Center, as a public pilot. Three engines transcribe colloquial Egyptian dialects, shift text between Arabic and 50 foreign languages, and voice Arabic script. The Arabic language AI tools sit at belmasry.aic.gov.eg.
Government Digital Delivery Agency — New Zealand's Government Digital Delivery Agency (GDDA) 2026 cross-agency AI survey counted 545 public sector AI use cases across 59 organisations, almost twice 2025's 272. Operational state-sector AI projects tripled to 167. On 18 August, the Leadership Development Centre and GDDA opened a free, self-paced AI Development Series for every public servant, teaching AI fundamentals plus safe, responsible use.
Ministry of Agro-Industry, Food Security, Blue Economy and Fisheries — Mauritius backs AI-driven precision farming via the Precision Farming and AI-Driven Food Production Scheme, paying half of project costs, capped at MUR 2 million per farmer. Robotics, IoT soil sensors, automated watering, vertical farms and precision agriculture technology qualify, part of a push for food security and higher harvests countrywide.
e-Albania (National Agency for the Information Society) — Albania has named four systems now driving AI in public administration through its e-Albania platform: the Legal Overlap System, which flags clashing or duplicated wording across legal acts; a module speeding translation and alignment of Albanian law with EU law inside the European Integration and Membership Platform; eIAD, which mines tax data for anomalies and risk; and Diella, which points citizens toward services and paperwork. Officials stress human staff still make the final call, though no launch dates were given for any of these four AI systems.
Lands Commission of Papua New Guinea — Papua New Guinea's Lands Commission sent twenty staff through AI training for land administration at the Institute of Technology and Innovation (ITI) in Port Moresby, held 17 to 20 August 2026. Modules covered Introduction to AI, AI for Office Productivity, and Data Visualization and Building Dashboards, building skills for land-dispute review and faster reporting. Chief Commissioner Molean Kilepak and the Commissioner for Operations joined the artificial intelligence and data course.
Ministerio de Ciencia, Tecnología, Conocimiento e Innovación — Chile's science ministry has put over 5,000 public officials through AI training via Stack Público and built TAM-GPT, a Latin American language model, with the National Centre for Artificial Intelligence. Six priorities for 2026-2030 target about 100 AI pilots in the state, some one million people trained and reconversion for 100,000 workers.
Government of the Bahamas — The Bahamas will put $3m of an Inter-American Development Bank loan into AI skills training, says Sebas Bastian, minister of innovation and national development. It splits between basic AI literacy for all and advanced tuition for Bahamians building technology, with courses announced shortly.
Office of the Minister of State for Digital Economy and Artificial Intelligence — Out of 4193 hopefuls, 155 young Libyans won places in an AI skills training scheme by Samsung Innovation Campus and Libya's Office of the Minister of State for Digital Economy and AI. Winners finish programming and AI instruction, then earn Samsung certification.
Tajikistan Ministry of Internal Affairs (trainees); Korea International Cooperation Agency (trainer) — Roughly 30 Tajik officials — from the Ministry of Internal Affairs, police and border guards — completed AI forensic investigation training at the Korea-Tajikistan Security Cooperation Seminar in Dushanbe, run by the Korea International Cooperation Agency. South Korea's National Police Agency and National Forensic Service covered deepfake and voice phishing detection, cryptocurrency tracing, and digital evidence from Telegram and large language models — one strand of AI-driven crime detection now reaching Central Asia. The programme ran five days, 3 to 7 August 2026.
Ministry of Construction — AI training government employees at Vietnam's Ministry of Construction reached over 1,000 officials on 4-5 August 2026 with CMC Technology Group across nearly 50 linked sites. Sessions covered management, policy drafting and paperwork, building AI skills public sector wide.
Ministère de la Santé du Burkina Faso — Burkina Faso's Ministry of Health delivered AI training for administrative secretaries to 72 headquarters and regional health directorate staff on 3–7 August 2026. Content spanned Microsoft Copilot, ChatGPT, advanced Excel and Word functions for drafting, filing and reporting, data security and AI ethics. The ministry's IT experts delivered the AI upskilling.
Société Nationale de Développement Informatique (SNDI) — SNDI delivered AI training for civil servants across 35 agencies in Côte d'Ivoire's Hambol region, handing certificates to 192 participants after a five-week course backed by the Ministry of Digital Transition and Technological Innovation. Curriculum topics ranged from artificial intelligence fundamentals through basic computing, web browsing and office applications, spanning 22 June to 29 July 2026. Katiola hosted the closing event marking completion, extending digital skills for government workers outside the capital.
Government Office of Estonia — The Government Office has opened a five million euro tender for AI skills training, seeking at least 100,000 working-age residents through one Eesti.ai workshop or learning bite during 2026 and 2027. Each participant takes two four-hour sessions, 75 percent plus face to face and offered in every county, open to workers from 18; some 65 percent of places go to existing AI users, 30 percent to beginners and five percent to the advanced. A three-month pilot ran April to June before this AI literacy programme, which sits under "Kõige AI-targem rahvas" and drew 5.2 of the 11 million euros Eesti.ai won from the June supplementary budget, after the Estonian government's January start.
Universidad Internacional de La Rioja — More than 400 civil servants across Ecuador gained grounding in AI ethics training for public servants through a programme run by Universidad Internacional de La Rioja (UNIR), touching on regulatory rules, openness and public governance. The curriculum unfolded at three separate venues - the National Assembly, the Council of the Judiciary and the Ministry of Defence - and paired instruction with 200 awarded scholarships toward a diploma focused on combating corruption. This AI governance training in the public sector explored how Ecuador's legal framework for artificial intelligence intersects with the ethical questions confronting government bodies.
Ministry of Finance of Sudan — Sudan's Ministry of Finance has begun AI training for civil servants, extending a new machine intelligence and digital transformation curriculum for public-sector roles to staff of the ministry plus the Audit Chamber, Internal Audit Department and Court Authority. Dr Mohamed Ali Jumaa launched the effort, framing it as a leap forward for institutions in the economic sector, with lessons oriented toward hands-on use across government work.
Government of Maldives — Digital for All, the Maldivian Government's AI literacy programme, targets 225,000 people by 2028 under the Maldives 2.0 agenda. The Ministry of Homeland Security and Ministry of Education jointly run the three-year scheme, offering free Dhivehi and English instruction via facilitators, island councils and online tools as part of wider digital skills training. Graduates earn nationally recognised digital competency badges.
Ministry of Scientific Research and Technological Innovation — The Republic of Congo's Ministry of Scientific Research and Technological Innovation delivered AI training in public administration to its staff. The session built practical skills in artificial intelligence for government work, part of a broader push toward AI skills for government staff.
Ministry of Communications and Information Technology (MCIT) and Egyptian Public Prosecution — Egypt has delivered AI training for prosecutors to over 2,000 Public Prosecution staff, covering responsible AI use, generative AI, project management and Microsoft 365 Copilot tools. The Ministry of Communications and Information Technology ran it with the National Training Academy, the Egyptian Anti-Corruption Academy, ITIDA and Microsoft, part of wider digital-skills building for legal personnel.
Instituto Tecnológico de las Américas (ITLA) / Policía Nacional — The Instituto Tecnológico de las Américas (ITLA) gave Dominican National Police leadership the country's first AI training for police in a five-week course, 'Artificial Intelligence Applied to Police Sciences', covering information analysis, decision-support tools, predictive analysis, automation and crime prevention. This law enforcement AI certification follows an ITLA-National Police memorandum of understanding.
Assemblée législative du Peuple du Burkina Faso — Roughly 30 lawmakers from Burkina Faso's National Assembly gained grounding in artificial intelligence during a joint session staged by the Legislative Assembly of the People with the Permanent Secretariat for Innovation and Forecasting on Emerging Digital Technologies (SPIVTEN) in Ouagadougou. Organisers designed the AI legislation training to sharpen deputies' capacity to comprehend, draft rules for and monitor emerging systems, an area officials have elevated to the 11th pillar within the nation's broader digital transformation agenda, closely tied to artificial intelligence oversight training goals.
Government of Honduras — Amazon Web Services and the Government of Honduras have launched AWS Entrena Honduras, an AI training programme giving 15,000 Hondurans no-cost artificial intelligence and cloud computing lessons. Sponsors also include the Central District Mayor's Office, the US Embassy and Grupo Financiero Atlántida, with distinct routes suited to ordinary learners, students and entrepreneurs from age 18. Finishers walk away holding a digital credential issued by AWS.
Marib Local Authority — Seba University and Marib governorate's local authority jointly trained 600 employees drawn from 30 state institutions in AI tools, split across six groups over five weeks. Sessions focused on deploying AI within daily office functions, aiming to lift public-service standards and speed digital transformation in Yemen.
Primature du Burkina Faso — Burkina Faso's Prime Minister's Office put senior managers through ethical AI training in Ouagadougou, under the banner 'Management tools for ethical and responsible AI'. Chief of staff Mamadou Dembélé launched the two-day workshop, expanding on instruction previously delivered to rank-and-file staff. Participants rehearsed crafting safe AI prompts that protect sensitive information and detecting bias within AI-generated results, reflecting a broader push toward responsible AI use in government.
Yemen Ministry of Youth and Sports — Yemen's Ministry of Youth and Sports has run AI training for female civil servants for a second time in Aden, drawing participants from ministries and state institutions nationwide. Minister Nayef al-Bakri patronised the four-day programme, which the Youth Welfare Fund supports and which covers digital innovation strategies alongside hands-on AI tool use for public administration. The push reflects wider efforts to widen women's participation in applying emerging technology within government work.
Labour Affairs Bureau of Macao — Macao's Labour Affairs Bureau has added AI courses for older residents to its Digital Skills Upgrading and Certification Programme, aiming to lift job prospects for this group. Over 40 sessions ran the year before, drawing 800-plus attendees. This AI skills programme for older workers extends the territory's digital-literacy push.
Ethiopian Artificial Intelligence Institute — Ethiopia's federal government finished an AI for government training programme on generative AI, prompt engineering and applied-AI tools. The Ethiopian Artificial Intelligence Institute led it with Addis Ababa University and the United Nations Development Programme under the AI UniPod framework, following government AI training already run in Dire Dawa and Sidama, with nationwide expansion planned.
Albanian Government (Prime Minister Edi Rama) — 140 senior civil servants across Albania's ministries have joined an AI training programme civil servants will rely on for sharper judgement and quicker decisions, Prime Minister Edi Rama said at Public Administration Week's third edition. This initial batch will grow into wider AI civil service training reaching other state bodies.
Ministério da Gestão e da Inovação em Serviços Públicos (MGI) — Brazil's Ministry of Management and Innovation (MGI) has delivered AI training for civil servants to over 167 thousand federal public servants through 11 distinct courses. Coursework runs via the Enap virtual school and Serpro, covering generative artificial intelligence, ethics, data protection, automated processes and data science as part of a wider drive to build AI skills for government workers. MGI has also introduced an AI Competency Matrix to steer further skills development.
Government of Guinea / Ministry Secretary General — Tamba Benoît Kamano, Guinea's Minister Secretary General of the Government, launched a Maferinya workshop teaching AI in public administration to Secretaries General across every ministry, backed by the UNDP. Over three days, officials explored artificial intelligence in governance to sharpen state performance, framed around responsible, effective public action.
Instituto de Inovação e Conhecimento (INIC) — The Institute of Innovation and Knowledge (INIC) put roughly forty officials from Sao Tome and Principe's public administration through a four-hour course on AI in public administration, taught by Dalila Rita, President of INIC's Board of Directors, at the Centro de Formação Profissional Brasil. Titled "Artificial Intelligence in Public Administration: Transforming the Future of Work and Innovation", it covered anti-corruption tools, citizen services, efficiency, data ownership and compatibility across government agencies.
Ministry of Economy, Finance and Budget (MEFB) — A Moroccan specialist guided officials through AI training in public finance at Conakry's Centre de Perfectionnement en Finances Publiques, commissioned by Guinea's Ministry of Economy, Finance and Budget. Attendees explored AI-driven administrative optimisation to pursue leaner workflows, cleaner oversight and stricter budget discipline nationwide.
Ministry of Communication, Technology and Innovation (MoCTI) — Sierra Leone's Ministry of Communication, Technology and Innovation and Ministry of Public Administration and Political Affairs will train 1,000 civil servants under a partnership agreement. The nine-month AI training for civil servants, via the Civil Service Training College, reaches 29 ministries, departments and agencies. Lessons span public sector AI skills, everyday digital tools, data management and cybersecurity under a wider digital transformation agenda.
Sierra Leone Police — A joint AI training for police pulled Sierra Leone Police staff nationwide together with Victory Teens Organisation instructors, covering cybercrime detection, digital probes and AI-driven policing tools. Deputy Inspector General Sahr Yomba Senesi wrapped things up, telling attendees to put lessons into action and pass them onward.
Papua New Guinea Centre for Judicial Excellence (PNGCJE) — 11 National Judicial Staff Services (NJSS) staff gained hands-on skills in Microsoft Copilot, Gemini, Claude, ChatGPT and NotebookLM via a judicial AI training programme at Waigani Supreme and National Courts. The Papua New Guinea Centre for Judicial Excellence (PNGCJE) funded the Generative AI for Beginners Bootcamp, pairing machine learning basics with generative AI architecture before tool practice. This judicial AI skills bootcamp ran three days, 23-25 June 2026.
Office for IT and eGovernment of Serbia — Serbia's Office for Information Technology and Electronic Administration teamed up with Microsoft to deliver AI-assisted public administration training to 1,500 civil servants using Microsoft Copilot. Lessons ranged from quicker document searches, summaries and drafting to handling routine bureaucratic duties, with government use of Microsoft Copilot extending to advanced modules in Word, Excel, PowerPoint and Outlook for select staff. The curriculum closed with building AI agents for workflow automation and instruction on responsible, data-protective AI practices.
Ministry of Communications and Information Technology (MCIT) and National Training Academy (NTA) — Egypt's Ministry of Communications and Information Technology and National Training Academy launched the Artificial Intelligence for Government Organisation Leaders National Initiative, offering AI training for public sector leaders. The hands-on senior-official AI upskilling covers cybersecurity, digital twins, agentic and generative AI, data governance, ethical AI use, and strategic and digital leadership. It follows the Digital Egypt Leaders National Initiative and serves the National AI Strategy 2025–2030.
Ministerio de Administración Pública (MAP) — Dominican Republic's Ministry of Public Administration (MAP) is funding 2,000 places on a course covering AI in public sector training for civil servants across all state entities, part of a growing government push in this area. The self-paced, fully remote generative AI course for civil servants runs through Academia BID with Inter-American Development Bank (IDB) backing, needs no prior background, and leads to a certificate covering generative AI applications and ethics in government work. Applications closed 4 August 2026; the full title is "Generative Artificial Intelligence: Fundamentals for the Public Function of Latin America and the Caribbean | DR Edition".
BMDS — Complex GDPR data requests now clear in hours instead of weeks, a 90% speed gain logged by Germany's Federal Ministry for Digitalisation and State Modernisation across Agentic AI Hub trials in 19 municipalities. Care benefit applications moved 45% faster and staff spent 78% less time chasing up meeting minutes, as the scheme piloted AI agents in local government spanning 20 administrative workflows. 70% of those pilot solutions, an early marker of public sector AI adoption, were already live in daily operation before the cut-off.
Government of Bangladesh and UNESCO — Dhaka hosted three days of AI training for civil servants under UNESCO and the Government of Bangladesh, pulling ministry staff across procurement and ICT roles. BRAC University and North South University taught within UNESCO's AI Ethics Experts Without Borders network, covering design, ethics and oversight. This builds public-sector AI capacity building nationwide.
Ministry of Post and Telecommunications — 993 people — ministry leaders and rank-and-file staff together — got through an AI training for civil servants course the Ministry of Post and Telecommunications titled Using Artificial Intelligence to Enhance Work Efficiency on the Journey Toward Digital Transformation, wrapped at a Phnom Penh event under Minister Chea Vandeth. Generative AI and large language models, he told the room, have stopped being distant scientific ideas and now reshape public-sector AI skilling and daily work alike.
Badan Kepegawaian Negara (BKN) — 145 thousand Aparatur Sipil Negara (ASN) staff across Indonesia stand to gain AI training for civil servants through a new alliance linking Badan Kepegawaian Negara (BKN) with Microsoft Indonesia, called GARUDA AI for Microsoft Elevate. Officials sealed the pact on 28 April 2026, spreading sessions from May through July 2026. Ahead of the public rollout, enrolment had already reached 12,551 people spread across six cohorts of AI for Public Impact and one cohort of AI Policy Lab for Leaders, part of a wider push on AI upskilling in government.
Polish Civil Service (Służba Cywilna) — On 10 June 2026, Poland's Civil Service launched a third free e-learning course on AI for government administration, hosted on the Civil Service e-learning system. Content spans core generative AI concepts, secure usage habits, crafting effective prompts and applying AI responsibly, forming part of Poland's wider push on AI in public administration for its staff. Two prior modules on the same platform came before it.
Ministry of Digital Affairs — Taiwan's Ministry of Digital Affairs and the Directorate-General of Personnel Administration issued the AI Civil Service Talent Identification Guidelines beta on AI training for civil servants. Staff fall into policy, application and development talent types with standard curricula and 40 course modules. For public-sector AI upskilling, dual-track certification directly accepts 25 industry certifications from AWS, Google, Microsoft and others.
Ministry of Education (Jordan) — Jordan's Ministry of Education counts over a million students and 100,000 teachers using Siraj, the AI learning assistant launched by the ministry in October 2025. Students put questions, get curriculum-tied explanations and revise for exams; teachers draft explanations and create curriculum-linked questions and activities, with integration into the ministry's learning management systems under way.
Ministry of Post and Telecommunications (MPTC) — Cambodia's Ministry of Post and Telecommunications delivered AI training for civil servants across 11 groups at CADT's Digital Government Institute, 28 May to 11 June 2026. Sessions covered AI basics, the national AI ecosystem, ethical risks, and tasks like document summaries, data analysis and content creation, building government-wide AI literacy.
Government of Jamaica — Jamaica backs GAINS (Growing AI Innovation and National Skills Programme) with JMD $545 million for AI skills certification across all 63 constituencies. A six-week AI Mastery Curriculum targets 1,260 HEART/NSTA-certified graduates per cohort, feeding a broader AI skills training effort. St Catherine and St Andrew pilot first, ahead of national rollout.
Belize Department of Civil Aviation — Staff at Belize's Department of Civil Aviation finished "Cleared for AI: A Civil Aviation Literacy Session," covering AI literacy in aviation. Topics included AI history, prompting, safety, ethics, and demos of FOD detection and RAG chatbots, aimed at building AI training for aviation staff while preserving human oversight.
Civil Service Training Institution — Fiji's Civil Service Training Institution in Nasese now folds AI civil service training into its coursework, Prime Minister Sitiveni Rabuka told Parliament, calling it part of wider AI-integrated government workforce development. Asked whether reforms covered AI, he confirmed the change, while conceding his own generation lagged furthest behind.
Ministry of Culture, Sports and Tourism — A four-day programme delivered by Vietnam's Ministry of Culture, Sports and Tourism instructed over 130 sector staff in AI in public administration. Instruction spanned the national AI strategy alongside applying artificial intelligence ethically inside government work, part of a wider digital transformation drive.
Judiciary of Eswatini — Eswatini's judiciary became Africa's first to run a pilot AI and electoral justice training for judges, with the Africa Electoral Justice Network. Sessions covered deepfakes, disinformation, digital evidence and remedies for election disputes tied to artificial intelligence, drawing on the AEJN Toolkit on AI, Elections and Electoral Justice.
Upravna akademija (Administrative Academy) of the Ministry of Public Administration — Over 40,000 civil servants in Slovenia completed AI training for civil servants under a scheme run by the Administrative Academy within the Ministry of Public Administration. More than 80 modules spanned generative tools for senior staff alongside routine departmental applications. The Republic of Slovenia and the European Union's NextGenerationEU recovery mechanism co-funded the effort, spreading AI literacy among public sector employees.
Digital Learning Hub and Luxembourg Ministry of State — Luxembourg's Elements of AI literacy training has reached over 7,600 residents since 2021, Minister Elisabeth Margue said, closing its 5th edition. That's roughly 1.1% of working-age adults, beating the 1% target for 2030 ahead of schedule. The free, nationwide AI literacy programme keeps growing.
Government of Malta — Malta became the site of OpenAI's debut national government tie-up, letting citizens and residents on its digital identity registry claim a full year of free ChatGPT Plus, the premium chatbot subscription. Unlocking the offer means finishing a no-cost AI literacy course co-designed with the University of Malta. The Malta Digital Innovation Authority administers the initiative.
Civil Service College Uganda — More than 1,000 government secretaries across Uganda attended AI training for government secretaries at the Civil Service College in Jinja City, run by the Association of Secretaries and Administrative Professionals in Uganda (ASAPU). Kathay Technologies led sessions covering productivity, reporting, data analysis and research using artificial intelligence tools for administrative work.
Judiciary of Seychelles — Judicial College (JUCOS), led by Supreme Court Judge Melchior Vidot, ran a two-day workshop for Seychelles' court staff on handling AI in judicial work responsibly. Judicial Officers, Judicial Research Councils and administrative personnel examined disclosure duties, accountability questions, ethical pitfalls and courtroom disputes tied to artificial intelligence in the justice system, building literacy around emerging tools. Organisers framed the sessions as groundwork for a Judiciary-wide strategy on managing AI going forward.
CANDI (private company) / Magistrates of Kankan — Magistrates and civil society figures in Kankan joined a 13 May 2026 workshop on AI in the judicial sector, run by Guinean firm CANDI under the banner 'Artificial Intelligence and Justice in Guinea'. First President Pierre Lamah, heading the Kankan Court of Appeal, praised the push, arguing that judges armed with artificial intelligence in justice work could produce rulings faster and lift overall court performance. Guinea's public administration was named as the broader target for such tools.
Tax Administration of Serbia — Tax inspectors in Serbia spent four days learning AI audit tools for tax administration at a European Union-backed session held 12-15 May 2026 under the IPA 2021 scheme. Markus Drost, the EU specialist running the course, showed how algorithms can surface high-risk audit candidates, though he insisted a person must sign off on every final assessment. Real case studies formed the basis of practical audit simulations, giving officials a chance to test AI-driven risk detection in tax administration hands-on.
Policía Nacional Civil (PNC) — Guatemala's National Civil Police (PNC) put 25 officers through a five-day AI for criminal analysis course with the UN Office on Drugs and Crime (UNODC), covering dark web methods, part of a push on artificial intelligence applied to crime analysis. SVET, the Public Prosecutor's Office and the National Institute for Strategic Security Studies also took part.
Office of the Head of the Civil Service of the Federation — Nigeria's Office of the Head of the Civil Service of the Federation has begun AI literacy training for civil servants: a three-day Programme in Abuja, backed by UNESCO and the European Union. Didi Walson-Jack calls it a lever for faster AI-driven change across the Federal Government. Trainers under UNESCO's ethical AI framework will carry AI skills training for government workers on to ministries and agencies.
Centro Nacional de Innovación y Tecnología Francisco El Chele Moreno (CI Nicaragua) / INATEC — CI Nicaragua, INATEC's Centro Nacional de Innovación y Tecnología Francisco El Chele Moreno, certified 110 public servants from MARENA and INPESCA in AI productivity tools after a six-week course. Staff practised using large language models and AI-integrated project management platforms to automate administrative work and improve service delivery. The course in AI-enabled work tools supports a state digital-skills drive to modernise public services.
XRİTDX and AI Academy — Government staff across Azerbaijan finished AI for product managers training from the Special Communications and Information Security Service (XRİTDX) and the AI Academy, learning to spot information security threats and craft AI-based tools. This product manager AI course backs the nation's cybersecurity strategy.
Government Office of the Czech Republic — Nine ministries sit among 50 public bodies across Czechia now signed onto the AI National Skilling Plan's 'AI na úřad' scheme, a collaboration led by the Government Office of the Czech Republic with Microsoft and Digiskills to build AI skills training for civil servants. First-year totals put ready-made learning environments at 24,000 workers, while 577 staff inside the Government Office itself wrapped up 45,153 online courses between them. The initiative pairs safe handling of Copilot Chat and comparable workplace software with wider AI literacy for government employees.
ACCIA (Association des cadres de contrôle, d'inspection et d'audit dans les structures publiques tunisiennes) — ACCIA, Tunisia's audit and inspection body for public staff, trained 160 OOUN officers on generative AI in public procurement in Tunis, under an ACCIA-OOUN partnership. Coverage spanned needs identification, tender writing, bid judging and compliance checks, reflecting growing interest in AI-assisted purchasing tools across Tunisian institutions.
Ministry of Vocational Training and Education / Ministry of Knowledge Economy, Startups and Micro-Enterprises — Algeria's Ministry of Vocational Training and Education and the Ministry of Knowledge Economy, Startups and Micro-Enterprises have started an AI training programme for ICT specialists at the El Rahmania National Specialized Vocational Training Institute in Algiers, targeting 500,000 workers by 2030 under this artificial intelligence skills scheme for technology workers. Instruction runs eight weeks, then four weeks building startup projects, across a 12-week course; instructor coaching began 15 January 2026. The effort ties to a wider digital strategy targeting artificial intelligence for 7% of GDP by 2027.
ASAN Service (State Agency for Citizen Services and Social Innovations) — ASAN Service, Azerbaijan's State Agency for Citizen Services and Social Innovations, is phasing in compulsory AI training for government employees, reaching 5,500 staff. The ASAN Innovative Development Center leads this public-sector AI upskilling, trialled in Baku, Shusha and Salyan, blending tool use, data handling, content generation and automation.
Judicial Education Institute of the Senior Courts of Belize — Judges and legal staff in Belize completed AI training for judges this April, run by the Judicial Education Institute of the Senior Courts. Chief Justice Louise Esther Blenman framed the workshop as part of modernising the judiciary, building capacity for artificial intelligence in the judiciary while preserving ethical standards.
Netcampus — Roughly 6,000 South African public-sector leaders have completed AI training for government officials delivered by Netcampus through Microsoft's Elevate programme. Participants spanned the Gauteng, Western Cape and Eastern Cape provincial administrations, the Department of Higher Education, TVET colleges, the South African Revenue Service, the University of Venda and the National School of Government. Netcampus combined virtual sessions, face-to-face workshops and a learning management system to deliver the artificial intelligence upskilling programme.
Ministry of Science and Technology (Mincyt) — Venezuela's Ministry of Science and Technology (Mincyt) is offering a free, self-guided online AI course, 'Introducción a la Inteligencia Artificial: Machine Learning y Deep Learning', through Prisma, its new educational platform. Anyone can enrol at prisma.mincyt.gob.ve; lessons span AI history, deep learning, machine learning algorithms and ethics. Minister Gabriela Jiménez Ramírez unveiled this AI training on World Creativity and Innovation Day.
Ministry of Economic Affairs, Entrepreneurship and Technological Innovation (EZOTI) — Staff at Suriname's Ministry of Economic Affairs, Entrepreneurship and Technological Innovation (EZOTI) received AI training for government staff, led by Abeleven Consulting. Sessions covered drafting text, condensing information, backing analysis and speeding paperwork, plus wider artificial intelligence workplace skills and responsible, purposeful use of the technology at work.
Yemen Customs Authority — Yemen Customs Authority (YCA) trained 17 staff in AI training for customs officials on 16 April 2026, closing a four-day course at its Aden centre with the Arab Academy's Institute of Administrative and Banking Training. Coursework covered AI tools in customs administration: analysing data, automating workflows and aiding decisions. YCA head Abdulhakim al-Qubati called it a strategic move toward a modernised customs system.
Ministère de la Transition Digitale du Burkina Faso — Burkina Faso's Ministry of Digital Transition delivered AI training for government workers to cabinet directors across all ministries and public institutions on 16 April 2026 in Ouagadougou. Minister Aminata Zerbo/Sabané and the President's chief of staff launched the single-day session, which covered core AI concepts, applying artificial intelligence within cabinet duties, sovereignty over digital systems, and managing associated risks. Officials frame the workshop, built around artificial intelligence instruction, as the opening step of a larger civil service training push.
Ministry of Public Service Efficiency and Digital Transformation (Togo AI Lab) — Togo AI Lab is training 260 people, including 60 public administration staff, through a six-week data science certification with Simplon from 13-14 April 2026. This machine learning training for civil servants covers Python and data processing. The Ministry of Public Service Efficiency and Digital Transformation organised this public-sector upskilling with GIZ.
Telecommunications Regulatory Board — Cameroon's Telecommunications Regulatory Board held a three-day expert-led seminar on AI in telecoms regulation for staff, 8–10 April 2026. Topics: everyday AI tools and AI-supported regulatory monitoring.
Teachers of Bhutan Volunteers / Ministry of Education and Skills Development — The DigiGyel Project began in Thimphu, a three-year push to carry AI literacy into Bhutan's high schools, with more than 150 remote schools first in line. More than 50,000 students and 300 teachers are to be taught AI skills, and more than 800 laptops go out nationwide, 260 reaching the education ministry at the start. The Switzerland-based NGO Dariu Foundation pays for the artificial intelligence training, run by Teachers of Bhutan Volunteers.
Ministry of Modernisation of the Administration and Public Service / ENA de Guinée — ENA de Guinée (École Nationale d'Administration) gave 35 hours of AI for public administration training via the FEMS+ tie-up with Quebec's ENAP. In Conakry, the module on Artificial Intelligence and administrative performance built 'manager-leaders' for Guinea's Simandou 2040 drive, covering generative AI in government administration, ethics, transparency and data protection.
Ministry of Communication, Digital Technology and Innovations — Ghana's Ministry of Communication, Digital Technology and Innovations, with UNESCO, is delivering AI literacy training for civil servants nationwide, an AI skills training for government staff effort built on a train-the-trainer model. An Accra-based group learns first, then passes the material to colleagues across ministries and agencies, one of four rounds finishing by May 2026. Machine learning, generative AI, algorithmic bias and data privacy make up the curriculum.
Ministry of ICT and Innovation (MINICT) — The Ministry of ICT and Innovation ran a multi-day Artificial Intelligence Trust and Safety Workshop on responsible AI in the judiciary, co-organised with UNDP Rwanda and supported by the Government of Canada. Specialists in finance, telecommunications, IT and cybersecurity worked on frameworks to keep judicial AI ethical and accountable.
Ministry of Communication, Digital Technology and Innovations — Ghana's Ministry of Communication, Digital Technology and Innovations has opened a second National AI Masterclass cohort in Akuse, part of its ongoing AI training for government officials, running to 25 March 2026. Deputy Directors and similar ranks build skills in AI, policy formulation and evidence-based decision-making. Around 100 officials across various ministries and agencies have already gone through this public sector AI capacity-building push.
Ministerio de Tecnologías de la Información y Comunicación (MITIC) — MITIC joined CONACYT, UNESCO and the EU to embed AI fundamentals for the public sector in "Fundamentos de la Inteligencia Artificial para el sector público", readying officials from multiple agencies for ethical, strategic tech use. A fresh cohort joined this government AI training in May 2026.
Ministry of ICT and Innovation (MINICT) — Rwanda's Ministry of ICT and Innovation ran AI training for chief digital officers drawn from government agencies nationwide. Cenfri delivered this public-sector AI upskilling, covering generative AI, practical use cases in government and the governance principles behind responsible deployment.
ASAN AI Hub — 100 officials became Azerbaijan's inaugural graduates of AI training for civil servants, run jointly by ASAN AI Hub and Apolitical across seven courses on prompt engineering, healthcare, procurement, climate policy and leadership. This government AI upskilling push drew broader demand: 500-plus sign-ups, 40-plus agencies.
Liberia Revenue Authority — Liberia Revenue Authority staff learned AI tools for tax communication at a three-day retreat held by its Communication, Media and Public Affairs Section. Employees from Taxpayer Services, Customs, Domestic Tax and other units trained in artificial intelligence for institutional communication, covering research, content creation and media monitoring.
Ministry of Communication, Technology and Innovation (MoCTI) — Sierra Leone's Ministry of Communication, Technology and Innovation delivered AI training for civil servants: over 50 senior leaders from ministries, departments and agencies. Qhala, the National Cybersecurity Coordination Centre and the Directorate of Science, Technology and Innovation co-ran the Miatta Conference Hall leadership course on cybersecurity, AI and emerging technologies. Aim: civil service AI skills so policy and strategic choices serve the public interest.
Ethiopian Human Rights Commission — Addis Ababa hosted a session where Ethiopian Human Rights Commission staff studied AI and human rights, examining how fast-growing AI affects rights protections nationwide. Trainers weighed AI's benefits against risks like bias, injustice and data protection breaches. EHRC folded this artificial intelligence rights protection work into its 2026-2030 strategic plan, listing further AI activities ahead.
Malawi Judiciary — Malawi's Judiciary Digitalisation Committee and Judiciary Training Committee ran a two-day workshop in Blantyre on responsible AI use in courts, backed by European Union funding through the Chilungamo II programme. Sessions covered how artificial intelligence functions and AI's role in judicial processes. Justice of Appeal Healey Potani stressed that judicial wisdom could never be replaced by AI, while Justice Fiona Mwale called for a clear institutional stance on where AI can responsibly assist.
Government of Malawi — Principal Secretaries from all Malawi cabinet ministries took one-day, high-level AI training for senior civil servants in Salima, covering service delivery, policy analysis, fraud detection and ethical adoption. Chief Secretary Dr Justin Saidi opened it, citing President Peter Mutharika's directive to speed up digitalising public services. UNDP backed the government AI upskilling via the Inclusive Digital Transformation for Malawi project; the World Bank, via the Digital Malawi Acceleration Project.
Ministry of Information Technology and Telecom — Pakistan's Ministry of Information Technology and Telecom is delivering AI training for government officials, covering 10,000 staff and aiming to lift digital literacy within public institutions. This forms part of the National AI Policy's capacity-building strand, running specialised AI workshops for public sector staff with industry partners.
National School of Government — AI fluency courses for public servants are now free and open nationwide under the NSG Digital Skills Programme 2026, run by South Africa's National School of Government, offering free AI training for government employees. Cybersecurity and mobile digital literacy training accompany part-time online study, delivered with the National Electronic Media Institute of South Africa and Microsoft's Elevate programme. Applications closed 31 July 2026.
Greek Ministry of Interior — Greece's Ministry of Interior renewed its Google cooperation deal, extending AI training for civil servants through December 2026 with Gemini AI built into the curriculum. Since 2023, 85,772 staff have finished digital skills courses, including 35,839 who joined AI-focused sessions across 77 webinars in 2024-2025, averaging 1,130 attendees each. The new phase targets hands-on AI fluency in daily public-sector work, broadening earlier government employee AI skills efforts.
Uganda Judiciary — Kampala's Judiciary headquarters hosted 80-plus research magistrates for a single day of Artificial Intelligence training on AI in legal research and judgment drafting. Acting Chief Registrar Pamela Lamunu Ocaya flagged the risk of leaning too heavily on AI, since it commonly invents cases that were never decided, and pressed for responsible AI use going forward.
Ministry of Digital Development of Kyrgyzstan — Kyrgyzstan's Ministry of Digital Development is opening AI training for civil servants via Apolitical, the world's largest online network for public-sector staff. This gives officials entry to the Google.org-backed Government AI Campus, aiming to train 1 million civil servants worldwide. It builds government AI literacy through specialised courses.
Ministry of Public Administration and Artificial Intelligence — A new six-month pilot called the AI Academy gives roughly 100 government workers in Trinidad and Tobago structured AI training for public servants, starting 3 February 2026. A group of 10 will go further, prototyping two to three AI applications inside a protected testing sandbox. Behind the scheme sits the Ministry of Public Administration and Artificial Intelligence, working alongside UNDP, UNICC and CAF to build AI skills development in government.
Rijksorganisatie voor Ontwikkeling, Digitalisering en Innovatie (RADIO) — Dutch civil servants must now finish AI training for civil servants, a 30-minute e-learning module capped by a quiz, built into the Basismodule Digitaal vakmanschap (Basic Digital Skills) programme. Content spans what AI actually is, situations where employees might run into it, and expectations for handling it responsibly across government. Officials will absorb this AI training for government workers into the compulsory Ambtelijk Vakmanschap (Civil Service Professionalism) curriculum going forward.
Ministry of Border Affairs, Myanmar — Myanmar's Ministry of Border Affairs launched AI training for government staff through its AI Technology Basic Applications Training Course. Union Minister for Border Affairs Lt-Gen Yar Pyae urged staff to apply AI to data analysis, calling it essential at work and home. The National Artificial Intelligence Development Project's Director outlined core AI skills for civil servants.
Rwanda Information Society Authority (RISA) — A half-year government AI training course, the Data Science Capacity Building Initiative (DSCBI), produced 41 graduates drawn from Rwandan public bodies, covering analytics and applied AI methods. RISA, NISR, AIMS Rwanda, the Mastercard Foundation and Cenfri built the scheme to sharpen evidence-based policymaking government-wide.
Guam Department of Labor — Almost every employee at Guam's Department of Labor has begun AI training for government staff, a partnership arrangement with the University of Guam. Built by UOG's Global Learning and Engagement unit alongside Pacific AI Media & Communications, the coursework launched with a module titled 'AI Literacy and Foundations', walking participants through fundamental ideas, available tools and rules for responsibly applying an artificial intelligence literacy programme inside government work. Two additional sessions remain, intended to guide workers toward folding AI into routine tasks.
Cambodia Academy of Digital Technology (CADT) — The Cambodia Academy of Digital Technology (CADT) and Phnom Penh Capital Administration gave AI training for civil servants to 518 staff. Three sessions on ethical AI use, prompt engineering, large language models and generative AI targeted work efficiency. Five civil service AI upskilling rounds have reached 1442.
Saudi Data and Artificial Intelligence Authority (SDAIA) — SDAIA and Saudi Arabia's Ministry of Education have unveiled SAMAI 2, extending AI training for government workforce staff into 11 ministries spanning healthcare, finance, justice and energy. The scheme follows SAMAI 1's 2025 reach of over 1.1 million trained citizens, now shifting toward sector-specific coursework for civil servants, aimed at lifting AI literacy among public-sector employees and productivity across government.
Brunei Economic Development Board — Brunei Economic Development Board has begun AI adoption for MSMEs through the AI for MSME Advancement in ASEAN programme: 550 firms by May 2027, in 11 cohorts. Small-business artificial intelligence training shows owners to write promotional posts automatically, answer buyers faster through AI-powered chatbots, and hold stock, sales and daily records together.
Ministry of Ethnic Affairs, Myanmar — Myanmar's Ministry of Ethnic Affairs held AI training for civil servants: an AI Technology Basic Practical Training Course in Nay Pyi Taw. Citing global use in education, health and industry, Union Minister for Border Affairs and Ethnic Affairs Lt-Gen Yar Pyae urged personnel to put AI upskilling to use, speeding departmental tasks.
Ministry of National Education and Vocational Training — Djibouti's Ministry of National Education and Vocational Training, the Organisation Internationale de la Francophonie and InnoVAI Djibouti certified 42 inspectors and pedagogical advisors in AI for education across seven workshops, October 2025 to January 2026. Output included sample lesson plans using generative AI tools, automated grading rubrics and a draft ethics charter for ministry-wide use. A CREFA/OIF centre ceremony marked completion.
Government of Madagascar — Madagascar's Government has opened ASAN'AI, a free 15-month AI digital skills training scheme for 1,300 young Malagasy. Delivered with the World Bank and private firms Intelcia, Konecta, Connecteo and Concentrix, this AI skills programme builds capacity for high-demand roles in business process outsourcing, including augmented customer experience work. A first batch of 280 trainees began placement from month three.
TalentCorp — TalentCorp and EY Consulting launched Jelajah AI MyMahir in Sungai Buloh, a RM110 million applied AI upskilling programme opened by Human Resources Minister Ramanan Ramakrishnan. Across 60 constituencies, hands-on AI learning reaches small business owners, entrepreneurs, women, workers and students; TalentCorp expects about 22,000 Malaysians, 12,000 via the roadshow and 10,000 via its other programmes.
PEZA / TESDA — The Philippine Economic Zone Authority has opened the country's first government-led AI training academy in Mandaue City, Cebu, with TESDA and private providers Advanced Institute of Technology and StackTrek. Workers at PEZA-registered firms, over 424 of them in Cebu, study applied AI for business, automation and generative AI tools for free.
Department of Education, Philippines — President Ferdinand Marcos Jr and Education Secretary Sonny Angara launched Project AGAP.AI, the Department of Education's AI training in schools on responsible AI use for teachers and students, at Quezon City Science High School. DepEd targets up to 1.5 million participants (1.05 million learners, 300,000 teachers and 150,000 parents) through AI Ready ASEAN Philippines training during 2026.
Ministry of Information Technology and Telecommunication — Pakistan's Civil Services Academy in Islamabad gave 150 probationary officers two-day intensive AI training for civil servants covering fundamentals, prompt engineering and ethical government use. Federal Minister for IT and Telecom Shaza Fatima Khawaja opened the Artificial Intelligence 101 module; training-of-trainers certified 30 faculty members as master trainers. The CSA curriculum formally includes officer AI instruction for all future batches.
The UK government's TechFirst programme commits £187 million to digital and AI skills training: TechYouth (£24m, 1 million secondary students over three years), TechGrad (£96.8m, 1,000 undergraduate scholarships a year, 100 Research MSc places, 100 elite AI scholarships), TechExpert (£48.4m, 500 domestic PhD students) and TechLocal (£18m, regional seed funding). Separately, industry and government aim to equip 7.5 million workers, roughly 20% of the workforce, with essential AI skills by 2030.
The EU's Digital Europe Programme has put over €294 million into digital skills funding for upskilling, reskilling and skilling the workforce. An EU-funded AI Skills Academy launches in 2026.
South Korea is spending ₩1.4 trillion (about US$960 million) on AI talent development in 2026: ₩900 billion for elementary and middle schools and ₩500 billion for high schools. AI-focused schools will expand from 730 to 2,000 by 2028, and vocational high-school departments offering AI courses from 20% to 50% by 2030.
Ministry of National Education of Türkiye — Türkiye's Ministry of National Education says 5,676,000 users signed into MEBİ, its AI learning platform for self-study, in the 2025-2026 academic year, finishing over seven million practice and assessment tests. Cumulatively, 4,092,992 students, 731,822 teachers and 852,069 parents have used it; AI-driven prompts show pupils what to work on, and teachers get tools to track academic progress.
INPS (National Social Security Institute) — Italy's national social security institute, INPS, lists 59 AI systems classified under AgID guidelines in its 2026 annual report, bringing AI in social security to a multi-agent conversational chatbot and certified-email routing. The routing system automatically directed 70.4% of 5.4 million messages analysed from April 2021 to April 2026 with 87% overall accuracy, and generative AI helps staff check reimbursement documents in the Estate Inpsieme welfare programme.
Communications Regulatory Agency of Bosnia and Herzegovina — More than 70 percent of staff at the Communications Regulatory Agency of Bosnia and Herzegovina reach for artificial intelligence tools regularly or occasionally, an internal poll found. AI adoption in public administration outpaces the rulebook: the regulator is still drafting one, and no statute covers such work. The Ministry of Foreign Affairs, the Central Election Commission and the Public Procurement Agency, whose chatbot fields queries on its eNabavke portal, report the same, marking AI use across state bodies.
Ministry of Transport, Telecommunications and Civil Aviation — Equatorial Guinea's Ministry of Transport, Telecommunications and Civil Aviation has begun AI and robotics training in Malabo alongside UNESCO. Instructors are being taught machine-learning software, 3D modelling and 3D printing before passing the skills to school-age learners. Between 2026 and 2028 the wider scheme aims to reach 1,000 youngsters and 400 trainers, delivered with the Ministry of Education; agriculture and healthcare are the first targets of this AI upskilling drive.
Centre National des Curricula (CNC) — Chad's Centre National des Curricula delivered AI training for education to roughly 60 curriculum designers, backed by UNESCO's International Bureau of Education. Held in N'Djamena, French sessions ran 10-12 August and Arabic sessions 13-15 August, covering summarising, curriculum review, assessment-grid design, translation and fact-checking. The work supports the 2025-2029 Projet de refondation de l'école tchadienne, building capacity for responsible AI use in curriculum design among national teaching-resource teams.
École nationale supérieure des travaux publics (ENSTP) — ENSTP (École nationale supérieure des travaux publics) ran generative AI training administrative staff sessions from 15 to 17 April 2026, focused on practical uses to lift productivity and efficiency. Its deputy director general said more rounds would follow, aiming to embed AI tools in administrative services long-term.
Government of Niger and United Nations System in Niger — One hundred (100) officials across Niger's public sector completed applied AI for crisis management coursework in Niamey during a two-week programme in June, organised jointly by the national government and the United Nations System with financial backing from the UN Trust Fund for Human Security. Representatives came from bodies including the High Authority for Peacebuilding, the Ministry of Interior and the National Institute of Statistics, gaining hands-on practice in artificial intelligence in emergency response through exercises in analysing datasets, compiling written reports and managing rapid-onset incidents. Branded ILLIMI Niger, the scheme sits within a broader push to build technical capacity inside government agencies.
United Nations Development Programme (with KOICA) — Senior Sao Tome and Principe officials completed AI training for civil servants, a two-day United Nations Development Programme course funded by the Korea International Cooperation Agency. Aims included digitalisation and AI capacity-building for government officials, sharpening how public services reach citizens, how decisions draw on data and how agencies answer for outcomes. A Director of Justice Policy and UNDP's Deputy Resident Representative opened proceedings.
Primature du Burkina Faso — Burkina Faso's Prime Minister's Office ran three-day AI training for public servants from its Ouagadougou central structures, 19 to 21 May 2026. Artificial intelligence's promise and dangers for state administration framed the course, in which volunteer Oumar Karaga, a Google Canada senior software engineer, taught civil service staff data analysis, AI fundamentals, digital writing tools and generative content's legal and ethical questions.
Academy of Management under the President of the Republic of Belarus — Civil servants across Belarus can now take a one-day generative AI training for civil servants at the Academy of Management under the President of the Republic of Belarus. The short course builds skill with ChatGPT, YandexGPT, Gemini and similar systems for producing text, images, audio and video, adding prompt techniques suited to state-body work. It reflects wider momentum behind AI adoption in public administration, with sign-ups running from January to July 2026.
Bundesakademie für öffentliche Verwaltung (BAköV) — The Federal Academy of Public Administration (BAköV) issued AI learning pathway (KI-Lernpfad) version 2.1, offering AI training for civil servants. Built with the AI Advisory Centre (BeKI), the pathway covers AI literacy basics, AI Act data protection and role-specific modules. Every German federal employee, not only AI or digitalisation officers, can join the civil service AI education.
Conseil National des Télécommunications (CONATEL) — Haiti's Conseil National des Télécommunications (CONATEL) ran generative AI training for staff over three days, 29-31 July. Sessions covered prompt engineering and practical tool use. École Supérieure d'Infotronique d'Haïti (ESIH) taught the course at CONATEL's Delmas site under Konbit IA, a push for building AI skills across public bodies.
Communications and Media Commission (CMC) — Iraq's Communications and Media Commission has launched AI training for government staff with Zain Iraq, the Irada Foundation and Kapita Research, running six months. Officials will study strategy, governance rules, tools and analytics, forming an AI governance skills programme inside the regulator. Baleegh Abu Kalal called it a first step.
Iraqi Judicial Institute — Iraq's Judicial Institute now treats artificial intelligence as a fixture of judge and prosecutor training, after UNESCO closed a five-month AI judicial training course. Brussels funded the work via the Social Media for Peace scheme. This AI literacy requirement for judicial education is now standard, not a single workshop.
National Assembly of Lao PDR — Laos's National Assembly delivered AI training for lawmakers to more than 150 members during a four-day session spanning all 18 electoral constituencies of the 10th legislature. The Rule of Law, Access to Justice and Good Governance (ROLAG) programme, run by LuxDev, supplied a senior adviser for an AI-driven era digital governance session covering parliamentary work. The Vice President opened the gathering, urging members to build skills suited to their constitutional duties.
General Authority for Communications and Informatics — Libya's General Authority for Communications and Informatics opened a year-long AI training for engineers and technicians, run with the Indian Embassy through the ITEC programme. Trainees take quantum computing and artificial intelligence upskilling at specialised centres in India.
CEPOL and EU4Security Moldova — Over 20 Moldovan officials finished a five-day AI investigations training course in Chișinău, run by CEPOL under the EU4Security Moldova project. Lessons spanned basics, data management, software tools, spotting patterns, and whether AI-derived evidence holds up in court, plus ethics and rights safeguards. It builds Moldova's capacity for AI in criminal investigations.
UNESCO Namibia — UNESCO Namibia trained 22 judges, magistrates and legal officers on Artificial Intelligence and judicial decision-making at a Windhoek workshop with the National Commission on Research, Science and Technology and the Konrad-Adenauer-Stiftung, running 23 to 25 June 2026. Topics included algorithmic transparency, machine-generated evidence, data safeguards and judicial independence, drawing on UNESCO's Global Toolkit on AI and the Rule of Law. Namibia's courts face growing exposure to AI in judicial processes.
Ministry of Energy, Water Resources and Water Supply, Gandaki Province — Gandaki Province's Ministry of Energy, Water Resources and Water Supply trained 28 officials in Pokhara through an AI training for government officials programme, drawn from water, irrigation and drinking water bodies across all 11 districts. They gained artificial intelligence skills for civil servants: reports, files, data and decisions.
Solomon Islands Government ICT Services (SIG ICTS) — Traseable Solutions' two-day AI for Software Engineers course trained eight teams of Solomon Islands Government (SIG) ICT Services (ICTS) officers in AI-assisted software development. Despite little or no coding background, the officers built working web and mobile prototypes for passenger declarations, asset management and customs duty calculation with AI coding tools including Claude Code.
Cabinet of the President of Suriname — Georgio Mosis of AethiQs Netherlands ran an AI for policymakers session at Suriname's eGov hall, held by the Cabinet of the President and the Ministry of Economic Affairs, Entrepreneurship and Technological Innovation (EZOTI). Attendees explored artificial intelligence in public administration for gains in efficiency, transparency and accountability, plus ethical guardrails and tools like ChatGPT and Notion AI.
Hof van Justitie (Court of Justice of Suriname) — Nearly all 28 judges of Suriname's Hof van Justitie received responsible AI use in courts training, alongside RAIOs, lawyers, notaries and legal writers. AI-Advies & Training led the work with the Centrum voor Democratie en Rechtspleging (CDR) and Simone Koole, casting AI in the judicial chain as support for human judgment, not its replacement.
Government of Turkmenistan — UNAPCICT/ESCAP, the Asian and Pacific Training Centre for Information and Communication Technology for Development, ran 'AI Essentials for Policymakers and Civil Servants', a two-day executive programme, in Ashgabat on 7-8 July. This AI training for civil servants gave senior officials governance, strategic oversight and risk-management skills for responsible AI across government and industry. Such public-sector AI upskilling comes while Turkmenistan works on its National AI Strategy.
Government of Madagascar (National Gendarmerie and Police) — Thirty Madagascar officers completed a regional course covering AI child protection tools under UNICRI's AI for Safer Children scheme, delivered alongside the UAE Ministry of Interior and the UN Office on Drugs and Crime. The Antananarivo programme, spread across three days, built skills in flagging abuse reports, connecting evidence trails and managing digital exhibits through AI-assisted investigation techniques, backed by practical exercises. A further 140-plus investigators from 16 additional nations followed remotely.
UAE Government — 97 percent AI tool use among government organisations defines this year's UAE government AI adoption milestone. Beyond 450,000 coders work nationwide, funded partly through a $1 billion pledge toward worldwide progress via AI initiatives. 44 percent of entities deploy high-performance computing spanning 91 use cases in healthcare, finance and security.
Civil Service Commission (CSC) — Kuwait's Civil Service Commission (CSC) launched Cloud Digital Leader, pairing AI training for government employees with cloud skills. Run with the Central Agency for Information Technology (CAIT) and Google Cloud, it is free, accredited, with a free licence for the certification exam, building artificial intelligence competencies.
Digitaal Vlaanderen (Flemish Government) — Belgium's Digitaal Vlaanderen built an AI literacy framework using five AI personas, showing local and Flemish civil servants which duties and skills each role needs. UNESCO co-developed the tool, which meets the EU AI Act's public-sector AI competence mandate and points staff toward linked training courses.
Ministry of Digital Transformation and Modernization of Administration — 300 Mauritanian civil servants are gaining AI and cybersecurity skills for public administration under a 12-week scheme from the Ministry of Digital Transformation and Modernization of Administration and Germany's GIZ. Nouakchott hosts the programme, branded 'Qualification du fonctionnaire', training staff from multiple sectors in using AI to boost output and safeguard data via the Centre de Formation et d'Échange à Distance — AI training for government workers at scale.
National Commission to Investigate Alleged Violations to Human Rights (NCIAVHR) — In Aden, 65 personnel from Yemen's National Commission to Investigate Alleged Violations to Human Rights (NCIAVHR) joined a UN-backed session covering AI for documenting human rights violations, arranged with the UN High Commissioner for Human Rights. Held 14 to 18 December 2025, topics ranged across OSINT methods, digital security, transitional justice and psychological wellbeing for rights workers. NCIAVHR framed it as advancing AI-assisted human rights documentation, building capacity for investigations that satisfy international standards.
San Marino Innovation / ULPA (Ufficio per il Lavoro e le Politiche Attive) — Civil servants at the Ufficio per il Lavoro e le Politiche Attive (ULPA) in San Marino launched a public sector AI training programme with San Marino Innovation, under a protocol signed 27 November 2025. Three hours covered large language models, automation tools like N8n, and low-code systems, showing staff how to simplify file handling and internal workflows. Participants also aimed to build software for public employment services, part of a drive toward AI automation for civil servants.
Ministry of Communications and Information Technology — MCIT and Scale AI launched ALIF, a government AI training scheme for Qatar's public workforce, delivered through the Qatar Digital Academy. Two strands make up the AI skilling programme: ALIF Core for office staff and ALIF Tech for specialists, both certified. The initiative debuted at World Summit AI Qatar 2025.
Ministry of Communications and Information Technology — Qatar's Ministry of Communications and Information Technology has nearly doubled Microsoft Copilot use for government AI adoption, expanding from 9 to 17 entities in phase two of the Adopt Microsoft Copilot programme. Phase one produced a 62% take-up rate, over 9,000 users, 1.7 million tasks completed and 240,000-plus hours saved, with public-sector AI uptake backed by Qatar Digital Academy training.
Ministry of Digital Development, Innovation and Communications of Mongolia — More than 60,000 Mongolian government workers gain free AI skills training for public servants under a new Oracle alliance with the Ministry of Digital Development, Innovation and Communications. Teachers, doctors, civil servants and the wider public gain three years of coursework via Oracle Academy and Oracle University, spanning artificial intelligence, databases, data handling, cybersecurity and IT. The effort ties to Mongolia's Digital First policy and National Big Data and AI Strategy, feeding a wider push toward AI literacy across the public workforce.
TNO — Dutch government AI applications climbed tenfold within twelve months, from just 8 in 2024 to 81 by 2025, per TNO's Overheidsbrede Monitor GenAI for the Ministry of the Interior. Of these 3 categories tracked, municipalities top the adoption list with 34 tools, largely chatbots sorting information, while 37 of 81 (46%) have moved into everyday routine use. For most public bodies, generative AI use in government still means backend support: 78% serve internal workflows rather than resident-facing tasks.
Government of Rwanda (Ministry of Public Service and Labour / MINICT) — Rwanda's Ministry of Public Service and Labour now compels AI training for civil servants across every government institution, whatever staff members' duties or technical grounding. A 31 December 2025 cutoff shifted to 15 January 2026 once logistical snags left some workers short of time. Skipping the mandatory AI courses for government employees brings disciplinary consequences under the public service framework.
Academy of Civil Service under the President of Turkmenistan — UNDP, Turkmenistan's Ministry of Communications and the Academy of Civil Service under the President of Turkmenistan trained 80 officers, IT specialists and university staff in AI in public administration, under the title Emerging Technologies and Artificial Intelligence (AI) in Public Administration. Estonia's E-Governance Academy supplied instructors; the wider push has now reached 120 trainees. Backers link this digital governance training for civil servants to the country's longer digital transformation goals.
State Agency for Public Service and Social Innovations under the President of Azerbaijan — Azerbaijan's State Agency for Public Service and Social Innovations has unveiled the ASAN AI Hub, an AI innovation government platform designed to unite scholars, officials, businesses and the public in one shared digital space, functioning as a government AI platform for public-private collaboration. Cash-prize contests posted there challenge undergraduates, academics, founders and civic organisations to devise fixes, positioning the hub as a crossroads linking universities, industry and the state. Its debut ceremony sealed partnerships with Apolitical, Breakthrough Social Enterprise and GenAI Works.
Government of Liberia — Ministry staff across Liberia gained AI training for government officials as the United Nations Development Programme partnered with Liberia's government on a nationwide modernisation drive during UNDP AI Week. Civil servants explored using artificial intelligence to sharpen service delivery and guide decisions, an early milestone toward public-sector AI capacity building.
Palestinian National School of Administration — The Palestinian National School of Administration delivered AI training for civil servants to 35 IT and human resources staff from 18 ministries and government institutions. 'AI in Government Work: Practical Applications and National Visions' targeted use cases in public administration with Arab Administrative Development Organisation, GOV Campus and International Advisory Group. Public-sector AI skills outputs will feed the State of Arab Government Administration 2026 report for Dubai's World Government Summit.
Banco Interamericano de Desarrollo (BID) with Agesic and ENAP — Uruguay's Inter-American Development Bank (BID), alongside Agesic and the National School of Public Administration (ENAP), has rolled out a generative AI training for public officials programme open to anyone working in government, whatever their function. Roughly seven hours long and hosted online through February 2026, the syllabus mixes basic AI theory with hands-on exercises, prompt engineering and risk evaluation. A digital certificate and BID Credential await those who complete it, reflecting a broader regional effort at AI skills training for civil servants.
Collectivité de Saint-Martin — The Collectivité de Saint-Martin held its first AI training for civil servants with the Centre National de la Fonction Publique Territoriale (CNFPT), targeting administrative managers. Eric Stattner, AI researcher at the Université des Antilles, led the Grand Case Beach Club workshop on how artificial intelligence might reshape public services including risk anticipation, data management and policy development. More public-sector AI upskilling sessions are planned.
Kenya School of Government — Kenya School of Government now offers free enrolment in the UNESCO-Oxford Massive Open Online Course on Artificial Intelligence and Digital Transformation in Government, under DigiKen. Officials at Job Grade CG8 and above get responsible AI training for civil servants, aiming for 20,000 trained by 2027 in this digital transformation training push.
Government of Zimbabwe — Zimbabwe will roll out mandatory AI training for civil servants, requiring at least 10,000 public employees to complete yearly coursework spanning artificial intelligence, cybersecurity, and data handling — an artificial intelligence coursework for public sector staff effort under a national digital push. Mthuli Ncube flagged Microsoft and Google certification as compulsory elements, housed within the Science, Technology, Digitalisation, Innovation and Human Capital Development cluster among next fiscal year's top reform goals.
Government of Mexico — An AI training centre in Mexico has opened with 10 mil Mexican students, launched by the Digital Transformation Agency and billed as the continent's biggest public AI and code school. Formally the Centro Público de Formación en Inteligencia Artificial, linked to 'Mexico: country of innovation', it offers 20 specialisms in AI, data, cloud, Java and cybersecurity, each with joint certification from Amazon, Google, IBM, Meta, Microsoft and Oracle. Organisers project 25 mil graduates a year.
Central Agency for Information Technology (CAIT) — Kuwait's Central Agency for Information Technology has joined Microsoft on AI training for government leaders, branded 'Enabling the Use of Artificial Intelligence for Ministers and Senior Officials in Government Institutions.' Copilot sessions launch the scheme. A Copilot Center of Excellence then extends AI training for public officials nationwide.
Sana'a Capital Secretariat (أمانة العاصمة) — Directors from every executive office under Sana'a Capital Secretariat gained skills in AI in administrative work at a course run by its Training and Qualification Office. Major General Abdul Ghani Jameel, Minister of State and Mayor of the Capital, launched the programme, centred on equipping senior officials with data analysis methods and practical AI uses to strengthen government services. Attendees also examined local barriers and possibilities for adopting such digital tools in public administration.
ILIA (Latin American AI Index) / CEPAL — AI adoption in Honduras scored 27.39 of 100 on the 2025 Latin American AI Index (ILIA) — 17th of 19 nations, third-last, ahead only of Bolivia and Venezuela. CEPAL and Chile's CENIA built the ranking and called Honduras a 'moderate adopter' despite wide gaps in connectivity, data centres and policymaking. Its score for governance of artificial intelligence, the vision and institutional capacity category, was zero.
Ministry of Justice and Supreme Judiciary Council — UNICRI joined Qatar's Supreme Judiciary Council to run a five-day session on AI in criminal justice for Court of Appeal and First Instance Court judges in The Hague. Qatar's Ministry of Justice, through its Centre for Legal and Judicial Studies, had already briefed legal professionals in two April sessions on criminal liability from AI technologies.
OSCE Academy in Bishkek — Bishkek hosted AI training for civil servants, mixing phishing defence with machine-learning skills for public administration across two days. National Institute for Strategic Initiatives representatives took part, exploring ChatGPT, DeepSeek and prompt-writing techniques. Attendees completing 'Digital Security and Artificial Intelligence for Sustainable Governance' walked away with certificates.
Government of Saint Lucia — AI training public sector staff in Saint Lucia sampled Recraft.ai, Pixverse.ai and ClipChamp at a Taiwanese Embassy workshop pairing image processing with generative software. College, school and government delegates practised directly rather than just watching demonstrations. Each left equipped to coach colleagues on using these tools for lessons, paperwork and promotion.
Royal Solomon Islands Police Force — The China Police Liaison Team (CPLT) gave AI anti-fraud training to more than 20 Royal Solomon Islands Police Force (RSIPF) officers at Honiara's Central police station. Lessons covered telecom fraud using intelligent script generation, video face-swapping and AI-powered voice synthesis, plus the 'three checks and three verifications' scam-detection method. Assistant Commissioner Patricia Leta credited the police scam-awareness session on AI with sharpening grasp of AI-related crimes and boosting operational readiness.
Zambia Evidence Lab (Ministry of Finance and National Planning) — A one-day AI for Government Workshop gave Zambian officials grounding in AI training for government officials, hosted by the Ministry of Finance and National Planning's embedded Zambia Evidence Lab (ZEL). Sessions ran from foundational artificial intelligence theory toward practical public-sector applications. The effort reflects a government AI literacy programme equipping policymakers to embed AI in their work.
Ministerio de Administración Pública (MAP) / PNUD — UNDP research shows 49.7% of Dominican Republic public servants use AI during work hours, marking notable AI adoption in public sector roles, with 67.1% relying on it more than once weekly. Among respondents, 33.6% use these tools for public digital services—health chatbot queries, complaint escalation, legal and administrative tasks. This reflects rising AI use among government employees nationwide.
Cabinet Office — One Big Thing 2025, the Cabinet Office's civil service AI training scheme "AI for All", gives every UK civil servant four 15-minute modules on artificial intelligence skills for government staff. It follows findings that just 28% felt confident using AI at work beforehand.
State School for Public Administration (Republic of Croatia) — Croatia's State School for Public Administration launched AI literacy training for civil servants, teaching staff to speed routine work with AI tools while keeping people accountable for checking outputs. Institute for Development and International Relations specialists ran the session, "Fundamentals of AI Literacy: Practical Application of AI Tools." Croatia joins the earliest nations giving government staff a formal AI skills curriculum.
Gobierno de Guanajuato — More than 8,000 public servants across Guanajuato completed the opening phase of generative AI training for public servants, run by the state's Innovation Institute alongside Microsoft, spanning 29 state agencies and all 46 municipalities. The focus was practical: applying AI-based decision-making tools for civil servants to daily work, from analysing data to improving how institutions handle routine processes.
Ministerio de Tecnologías de la Información y las Comunicaciones (MinTIC) — Free AI training for government workers launches in Colombia as MinTIC and the University of Cartagena fill 3,000 places on an 80-hour virtual diploma starting 6 October. Minister Carina Murcia called it central to making artificial intelligence a genuine driver of efficiency, transparency and better citizen services. The scheme ties into CONPES 4144, the country's broader AI policy framework, and reflects a semantic push toward wider public-sector adoption of an artificial intelligence skills programme for public servants.
Ministry of the Interior and Safety — By 2030, 2 myung (20,000) civil servants will hold "AI champion" status under a new South Korean certification scheme spanning Green, Blue, and Black tiers, designed by the Ministry of the Interior and Safety to embed AI in government work. Demand emerged early: a 220-seat pilot logged 770 applications. Public institutions now get shared access to the ministry's AI literacy curriculum, and completion figures feed directly into staff performance reviews.
Ministry of Justice of Cabo Verde — 226 legal staff nationwide began AI training for justice professionals in Cabo Verde, an AI training programme for the judiciary organised by IMIJ under the Ministry of Justice. Concluding 3 October, sessions cover core AI concepts and practical courtroom applications, backed by UNDP alongside the Directorate-General for Justice Policy. Delegates arrived from every island.
KPMG in Canada — AI adoption in government stays low across Canada: KPMG in Canada polled 349 Canadian public servants, finding just 22 per cent of public sector bodies have formally embraced AI. Yet 48 per cent of staff use it unofficially, half via open tools such as ChatGPT instead of vetted systems — fuelling accuracy, privacy and security exposure. Weak literacy compounds this: 85 per cent call public sector AI knowledge poor, and 69 per cent confirm real unease over generative AI tools in the public sector.
Egyptian Ministry of Communications and Information Technology and Ministry of Justice — Egypt has certified 462 legal professionals through the 'AI Ambassadors - State Cases Authority Batch' scheme, jointly run by the Ministry of Communications and Information Technology and the Ministry of Justice. The National Telecommunication Institute taught the 36-hour course on AI ethics, legislative challenges, and tools such as ChatGPT and OpenAI, reflecting broader AI in legal training efforts. Graduates across all batches of this AI for legal professionals scheme now exceed 1,100 since it began.
Ministry of Industry, Science, Technology and Innovation (MISTI) — MISTI trained roughly 400 staff in AI for public sector innovation. The Ministry of Industry, Science, Technology and Innovation session built skills in applying artificial intelligence in government operations, covering core concepts, ethics and hands-on public-service use cases, while pushing a proactive, adoption-ready outlook among attendees.
Senado de Puerto Rico — Custom GPT models were the focus for staff at the Senado de Puerto Rico, who each built their own tailored chatbot in an AI training for legislative staff held 27 August 2025. Senate President Thomas Rivera Schatz backed the session, coordinated by Vice President Carmelo Ríos Santiago, titled 'Inteligencia Artificial más Inteligencia Emocional, crea tu GPT personalizado'. Employees practised generative AI for drafting emails, running social media, and making images and audio, reflecting broader interest in AI in legislative offices.
Agency for Innovation and Digital Technologies under the President of Tajikistan — Bonoor, a free AI assistant government services tool replying only in Tajik, sits inside ehukumat.tj, Tajikistan's new digital gateway from the Agency for Innovation and Digital Technologies. Since its March 2025 pilot, over 1.5 million residents joined, drawing on 20 online services spanning 45 state bodies. ChatGPT falters on Tajik syntax, so this AI-driven public service delivery model closes that gap.
Government of Chad — Chad's government enrolled 2,000 young citizens aged 15 to 35 in a nationwide AI training programme, run jointly by the Ministry of Telecommunications, the Ministry of Higher Education and the UNESCO ACCES-TIC Chair. Sessions spanned 4 to 14 August 2025 and mixed foundational machine learning ideas with hands-on uses spanning farming, healthcare, schooling and broader economic sectors. Organisers closed the effort at the Ministry of Foreign Affairs by handing certificates to graduates, naming 41 standout participants for distinction as part of this AI skills training push.
Ministry of Education, ICESCO — ICESCO held a three-day Tunis workshop on AI in education and strategic foresight for Tunisia's Ministry of Education. The session ran with the Tunisian National Commission for Education, Science and Culture and the National Centre for Educational Technologies, part of wider regional AI policy in education efforts.
Government of Ghana (Ministry of Communication, Digital Technology and Innovations with UNDP) — Ghana's Ministry of Communication, Digital Technology and Innovations and UNDP ran a three-day Ministerial AI Bootcamp in July 2025, an AI capacity building government drive for Cabinet Ministers, Ministers of State and focal persons. Chief of Staff Julius Debrah opened it, tying AI adoption metrics to ministerial KPIs with yearly reporting required.
Muscat Governorate — A week-long course for civil servants in Oman's Muscat Governorate, "Implementing Artificial Intelligence in the Work Environment," ran July 20 to 24, 2025, delivering AI training for government employees. Instruction blended foundational theory, applied techniques and real-world examples of AI-assisted decision-making, meant to leave staff capable of building AI-based tools themselves. The underlying purpose was boosting how efficiently public services operate.
Institute of Public Administration — Government AI adoption in Bulgaria scored 49 out of 100 in the country's first AI Readiness Index, published by the Institute of Public Administration. The public sector AI readiness rating, drawn from polling 6,254 civil servants in April and May 2025, places the administration in a 'building capacity' phase. Openness and learning motivation stood out as strengths, while strategy, leadership and AI Act preparedness lagged.
Digital Government Development Agency (DGA) and Microsoft Thailand — Roughly 2,000 officials from agencies across Thailand completed AI training for government officers, run jointly by Microsoft Thailand and the Digital Government Development Agency (DGA) as 'Tech for Gov 2025: NextGen Gov AI Training Program'. Topics ranged from core machine-learning concepts to ethical rollout practices, building AI skills in government roles, and ending in a hackathon where teams pitched their own AI builds.
Ministry of Digital Technologies of Uzbekistan — Uzbekistan's One Million AI Leaders programme, run with the UAE, targets 100,000 young people this year via its AI study platform. The free AI training programme covers five tracks — AI fundamentals, generative AI, prompt engineering, ethical AI use — open to students, entrepreneurs and 100 tech enthusiasts, part of a wider AI skills initiative.
Ministry of Digital Transition and Administrative Reform — An AI supervisors training programme Morocco has launched covers 770 recruits from Guelmim, Tadla, Agadir and other regions, who will run AI awareness sessions for young people at youth centres nationwide. Minister Amal El Fallah Seghrouchni's Ministry of Digital Transition partners with the Ministry of Youth, Culture and Communication and Mohammed VI Polytechnic University for balanced coverage.
Czech government and Microsoft — Czechia aims to give more than 350,000 residents artificial intelligence training within a year through the AI National Skilling Plan, a joint venture between the Czech government and Microsoft. Backing comes from Microsoft, which is contributing over CZK 10 million in first-phase funds, reaching civil servants, teachers, students, judges and jobseekers under this AI training in government drive. Linked to the National AI Strategy 2030, the scheme advances the nation's broader AI skills development goals.
Banque de la République d'Haïti (BRH) — BRH, the Bank of the Republic of Haiti, now delivers AI skills in central banking coaching to managers, spanning artificial intelligence, cybersecurity and ethics. Governor Ronald Gabriel unveiled the plan at the Ayiti IA 2025 summit as a route to embed innovation within institutional culture, and a possible template for national tech upskilling policy.
State Commission on the State Language under the President of Kyrgyzstan — Kyrgyzstan's State Commission on the State Language, with UlutSoft, debuts Aitil, the nation's first Kyrgyz-language AI platform, running solely in Kyrgyz. It bundles speech recognition, text-to-audio, a chatbot, OCR and translation across 10 languages, backing the tongue digitally via Kyrgyz AI translation.
Trinidad and Tobago Police Service (TTPS) — The Trinidad and Tobago Police Service (TTPS) Tobago Division trained 20 officers in a one-day pilot, 'AI For Law Enforcement', at Tobago Information Technology Limited, Signal Hill. This AI literacy training for police, run with GNC Labs and sponsored by Jade Monkey Bar and Casino, covered spotting deepfakes and applying AI tools. Policing AI upskilling sits within the division's operations plan.
Government of Mauritius — Rs 25 million will equip Mauritius government departments with AI tools under the Public Sector AI Programme, run by a new AI Unit inside the Ministry of Information Technology, Communication and Innovation. This AI in public ministries push sits within Budget 2025-2026, which also funds an AI Innovation Start-Up Programme and tax deductions up to Rs 150,000 for start-ups and MSMEs investing in artificial intelligence in government. The aim is sharper policymaking and service delivery nationwide.
Ministry of Digital Development and Agency for Civil Service Affairs of Kazakhstan — Kazakhstan's Ministry of Digital Development, Innovation and Aerospace Industry, with the Agency for Civil Service Affairs, rolled out AI Qyzmet: AI training for civil servants using five-minute videos. Officials learn to apply artificial intelligence in planning, data handling, decisions and automating tasks. This Central Eurasia first targets 30,000 public employees yearly, an AI skills training for government workers push.
Ministry of Digital Transformation of Ukraine — A new scheme gives Ukraine's public workforce, from senior officials to IT staff, access to AI training for civil servants, backed jointly by the Ministry of Digital Transformation, Google and CDTO Campus. Under the banner Artificial Intelligence for the Public Sector, the free online course builds skills in responsible AI adoption in the public sector, automating repetitive work, and writing sharper prompts. The Ministry of Economy was the pilot agency, kicking off its cohort in June 2025.
Singapore Government — About 50,000 Singapore public officers, a third of the total, now regularly use the government AI chatbot Pair for research and writing tasks. Josephine Teo said active users have doubled since 2024, with roughly 100,000 officers having tried it. Staff have also built thousands of bots on the AIBots platform as an AI chatbot for public officers, covering HR queries and budget guidance.
Office of the Privacy Commissioner for Personal Data (PCPD) — Hong Kong's Privacy Commissioner for Personal Data checked 60 organisations, finding AI adoption in Hong Kong businesses at 80%, up 5% on 2024. Over half, 54%, ran three-plus AI systems in daily operations, mainly client support, marketing and back-office tasks. No Personal Data Ordinance breaches turned up.
Krajowa Szkoła Administracji Publicznej (KSAP) — KSAP trained Polish civil servants in AI in public administration through a 2025 series: prompt engineering, drafting with ChatGPT and Microsoft Copilot, ethical artificial intelligence use in government. "AI in the Office" opened the run as one online day, plus sessions on 30 June, 8 October and 13 October.
Government of Bermuda — Bermuda now ranks first worldwide, its government launching the Centre for Finance, Technology and Entrepreneurship's Supercharged Programme for Bermudians. Building atop the earlier Fintech Training Programme, this feeds a wider push preparing citizens for an AI-driven economy through national AI workforce training, anchored in AI skills training as a top priority.
Information & eGovernment Authority (iGA) — Bahrain's iGA taught staff across its directorates AI tools for government productivity, running a workshop titled 'AI Tools for Employee Productivity' at the Bahrain Institute of Banking and Finance. Its Innovation and Advanced Technologies Directorate led the session, covering prompt engineering and generative AI for civil servants alongside hands-on daily-task exercises.
Ministry of Justice / UNESCO / GIZ — UNESCO, GIZ and Rwanda's Ministry of Justice ran a Kigali workshop introducing 31 judicial staff to AI in judicial systems. Attendees explored algorithmic influence over court rulings and case handling within data protection standards. Concerns raised covered bias and opacity affecting AI-assisted justice delivery.
Tamkeen (Labour Fund) — Tamkeen's AI training programme aims to give 50,000 Bahrainis artificial intelligence expertise by 2030. Bahrain's Labour Fund split the effort into three strands: executives get grounding in adopting AI strategically, ordinary staff learn automation and analytics tools day-to-day, and specialists build systems from scratch. Outside firms and domestic institutions jointly run the coursework, ordered into motion by the Chairman of Tamkeen's Board of Directors.
Departamento de Desarrollo Económico y Comercio (DDEC) — Over 120 civil servants from eight Puerto Rico agencies received AI training for government employees during the week of 7 April 2025, via 'Inteligencia Artificial para el Gobierno' (AI for Government). Microsoft and PRITS backed the Departamento de Desarrollo Económico y Comercio (DDEC), which assembled staff from Education, Health, Family, Treasury, Public Security, Housing, Justice, the Planning Board, Budget Office, Medical Services Administration and Police. Coursework covered machine-learning basics, public-agency applications and hands-on tool practice — an AI for Government course others might replicate.
Presidency Digital Transformation Office — Government AI training for civil servants in Turkish public administration reached its first milestone as the Presidency Digital Transformation Office finished a 140-hour course for 24 technical staff chosen by the Ministry of National Defence, under the Public Sector AI Transformation Project. Graduates covering machine learning, deep learning and open-source tools received certificates, completing this public sector AI transformation programme's opening cohort.
Government of New Caledonia — New Caledonia's government and the Chamber of Commerce and Industry (CCI) gave nearly 100 agency heads AI training for public administration in a single day. Sessions covered core concepts, prompt engineering, intelligent agents and cases like automating workflows, virtual assistants and predictive analytics, plus adoption risks. The push reflects wider generative AI adoption in government across the territory.
Instituto Nacional de Administración Pública (INAP) — Civil servants in Guatemala can now enrol in AI training for civil servants through Guatemala's National Institute of Public Administration (INAP), one of 49 fresh entries in its catalogue spanning diplomas, postgraduate degrees and shorter programmes. Delivered entirely online across 48 hours over two months, the AI in public administration course walks learners through core concepts, available tools, ethical questions and hands-on applications intended to sharpen government transparency and efficiency. Enrolment starts 17 March 2025 ahead of a May launch.
Escuela Superior de Innovación en la Administración Pública (ESIAP) — 21 employees drawn from nine Salvadoran public bodies now hold Google Cloud Platform credentials, a milestone for AI adoption in government. ESIAP rector and Vice-President Félix Ulloa presented the honours once trainees wrapped a Google Day course on cloud configuration. Backers described the initiative as advancing digital transformation of public services nationwide.
Ministry of Digital / National AI Office (NAIO) with Google Cloud — Malaysia's Ministry of Digital extended AI in public sector access to 445,000 civil servants via AI at Work 2.0, built with NAIO and Google Cloud, placing Gemini inside Google Workspace. Sessions teach report writing, information review and lighter workloads through government AI adoption.
Academy of Civil Service under the President of Turkmenistan — A four-day course paired the Academy of Civil Service under the President of Turkmenistan with UNDP in Ashgabat, briefing officials on AI in government administration. Lessons cast artificial intelligence as fuel for digitalisation, touching its presence in official platforms and its lift to civic growth. Viktor Dostov of RUDN University steered the block on AI in the public sector, walking staff through fresh digital tools.
Law enforcement was the largest AI use category at the US Department of Homeland Security in 2025: 86 of 238 reported AI use cases, including 49 at Customs and Border Protection and 29 at Immigration and Customs Enforcement. Of those, 35 were classed high-impact.
Government of Mauritius — Mauritius's Government unveiled AI for All on 29 October in Port Louis, driving AI literacy Mauritius-wide with UN backing through the Ministry of Information Technology, Communication and Innovation. DIVA, the Digital Interactive Virtual Assistant, anchors this push for public AI awareness, explaining machine intelligence in English, French and Creole. Copies go to schools and social centres under the Digital Transformation Blueprint 2025-2029.
Government of the Russian Federation — Russia's Ministry of Labour has enrolled over 450 federal officials in AI training for civil servants, part of a government AI skills programme rolled out nationwide. Instruction blends classroom sessions with remote-access tools, and successful participants earn a professional development certificate. Deputy Prime Minister Dmitry Grigorenko tied the effort to a simple aim: without trained staff, AI adoption across the public sector stalls.
Government of Sri Lanka (Ministry of Digital Economy, ICTA) — About 1,000 executive officers from ministries across Sri Lanka joined AI training for public officials at Temple Trees, Colombo, run jointly by the Ministry of Digital Economy and the Information and Communication Technology Agency (ICTA). The single large workshop covered generative tools including ChatGPT and Gemini, aiming to lift digital literacy across government ranks.
Ministry of Public Utilities, Energy & Logistics (Belize) — 34 officials, including Minister Michel Chebat and Dr. Leroy Almendarez, gained government leadership training via a two-day workshop jointly run by Belize's Ministry of Public Utilities, Energy & Logistics and the Belize Port Authority. Participants built hands-on capability for steering AI responsibly inside public bodies.
Conseil d'Etat de la République Démocratique du Congo — High-level magistrates at the Democratic Republic of the Congo's Conseil d'Etat (State Council) gained generative AI legal practice skills over two days, 5-6 May 2025, advancing AI in the legal profession locally. First president Brigitte Nsensele led, taught by Ferdinand Vaweka, with IT kit given to attendees.
Ministry of Digital Transformation, Innovation and Modernization of Administration (MTNIMA) — Mauritania's Ministry of Digital Transformation, Innovation and Modernization of Administration trained 30-plus officials from the Justice, Transport, Finance and Digital Transformation ministries on an AI-powered low-code platform for government services in Nouakchott. Kitsoft, a Ukrainian GovTech firm, and the ITU taught staff to prototype and deploy eID-linked tools via Liquio — AI-assisted government software development in practice. The work supports a wider ITU-led Global GovStack scheme for independent digital service-building.
Ministry of ICT and Innovation (MINICT) — Rwanda's Ministry of ICT and Innovation, with GIZ, Women Political Leaders, Cenfri and the Certa Foundation, gave AI training to about 150 Parliament members from the Chamber of Deputies, Senate and Office of the Clerk in one day. Lessons built AI literacy via basics, generative AI, governance, inclusive AI, and legislative tools for document analysis, translation and budget work.
Government of Togo — Roughly 60 civil servants in Togo gained hands-on exposure to ChatGPT, Copilot, Gemini and Claude through UNDP's four-day AI training for civil servants. The curriculum spanned drafting official documents, data analysis and strategic planning, plus algorithmic bias and data protection — themes central to generative AI in public administration.
Agesic (with PNUD/UNDP) — Uruguay's Agesic and PNUD, the UN Development Programme, delivered AI training for government to public servants across national ministries over three days. Staff covered core concepts, ethical concerns and administration examples, then practised designing artificial intelligence for public administration tools directly. The work supports the country's wider national data and AI strategy.
The National Audit Office measured government AI adoption across 87 UK central government bodies for its March 2024 report. By autumn 2023, 37% had deployed AI and 70% were piloting or planning it.
The US federal government's 2025 AI use case inventory lists 3,611 individually-reported AI use cases across 56 agency submissions, 445 of them high-impact. The Department of Justice reported 314 (114 high-impact) and the Department of Homeland Security 238 (55 high-impact).
In the 2023 OECD Digital Government Index, about 70% of participating countries had used AI to improve internal government operations but only 30% for policymaking. Back-office use is far ahead of policy work.
Sharjah Judicial Council — Sharjah Judicial Council adopted a court AI guide, the latest Sharjah AI regulation. The guide covers human oversight, cybersecurity, data protection, risk management and use cases across judicial and administrative work.
Ministry of the Interior and Safety (South Korea) — Ministry of the Interior and Safety confirmed public-sector AI ethics standards with 68 self-check items, a pillar of South Korea AI policy. Six values anchor them: privacy, safety, accountability, transparency, equity and public interest. People keep final responsibility over AI-supported administrative decisions, with named owners and human review.
Government of South Australia — South Australia AI policy moved as Dr Iain Ross AO took charge of the Royal Commission into AI alongside Kate Pounder and Professor Christopher Manning, starting immediately. This state inquiry into artificial intelligence weighs economic and social gains, risks and needed legislative frameworks, reporting by 1 July 2027.
Supreme Court of Finland; Åland Lagting — Åland AI regulation cleared review as Finland's Supreme Court held the Regional Act on Supervision of Artificial Intelligence within the islands' powers, with no bar to commencement. Passed by the Lagting on 17 June 2026 alongside amendments to sections 1 and 4 of the regional act on market control of certain products, this AI oversight law supplements EU Regulation 2024/1689.
Ministry of Digital Economy, Digitalization and Innovation (Gabon) — The Ministry of Digital Economy, Digitalization and Innovation opened drafting on Gabon AI policy at the Cité de la Démocratie in Libreville, steered by the United Nations Development Programme with the AILA method. For three days, officials, watchdogs, operators, academics, businesses and community groups weighed training, data, networks, oversight, cyber risk and rights. Private talks come next; no date to adopt the artificial intelligence plan has been published.
Ministry of Digital Development of Kyrgyzstan — A public-comment draft of Kyrgyzstan AI regulation from the Ministry of Digital Development sets ethics rules for algorithm design and use. Machine-generated content needs labels, users must be told when an AI agent replies, and government systems require safety review before launch. Kyrgyz-language function is compulsory for state tools, excluding military and police uses from these AI governance rules.
Asamblea Legislativa de El Salvador — The Legislative Assembly passed, by 57 votes, an El Salvador AI regulation giving the National Artificial Intelligence, Robotics and Similar Technologies Projects led or coordinated by the National Artificial Intelligence Agency (ANIA) a separate purchasing regime. Declared of public utility, those projects can buy anything from works and licences to data, models, compute and infrastructure under AI procurement rules set apart from the Public Procurement Law.
City of Centralia, Missouri — The City of Centralia, Missouri, has adopted a generative AI policy with a sign-off form stating when staff may use AI. Employees need supervisor approval to apply AI to city business and remain answerable for checking all output, under the new AI rules for staff.
Government of Lao PDR — The Lao cabinet agreed in principle on the draft National Artificial Intelligence Strategy for 2026-2035 at a 22-23 September session under Prime Minister Sonexay Siphandone, advancing Laos AI policy. The Ministry of Technology and Communications must revise the national AI plan towards infrastructure, databases, skilled staff and AI's highest-value uses.
Egyptian Center for Responsible AI (ECRAI) and National Council for Artificial Intelligence — Egypt's National Council for Artificial Intelligence, known as NCAI, weighed the draft Guidelines for the Procurement of AI Systems by Government Entities during its September 2026 session, pairing it with a separate plan for responsible generative AI use across public bodies. Behind the draft sits work by the Egyptian Center for Responsible AI, which gathered input from state agencies and private companies to fix technical thresholds, risk checks and supplier vetting tied to governance rules for AI procurement in government. The goal: a hands-on reference letting agencies handle public sector AI purchasing and judge vendors on consistent terms.
Revolutionary Government of Zanzibar — Zanzibar's Revolutionary Government unveiled its AI and Digital Policy, a framework for the archipelago's digital growth and artificial intelligence uptake. Officials presented this AI governance framework at the Tech & AI International Expo 2026, organised by Africa Business Inc with Zanzibar's Ministry of Communications, Information Technology and Innovation.
Communications and Media Commission (CMC) — A 30-day feedback period has opened on Iraq's first AI regulation Iraq draft, a regulatory framework for AI services from the Communications and Media Commission. Rooted in Legislative Order No. 65 of 2004, it targets oversight of AI development and deployment, tighter privacy, sharper openness, digital sovereignty, and industry investment.
National Commission on Research, Science and Technology (NCRST) — The National Commission on Research, Science and Technology (NCRST) created the National Artificial Intelligence Institute under Cabinet Decision No. 14th/23.08.22/001, advancing Namibia AI policy. Four units (Development, Datasets, AI Innovation, Research and Development) steer this national AI hub's studies, ideas, data and homegrown tool rollout for ministries and firms.
Government of Spain — Spain AI policy now follows the IA360 Plan, a 12-month programme of concrete steps Pedro Sánchez unveiled to capture economic gains while putting safety, trust and protection of children and other vulnerable groups first. Measures span compute and new models, cybersecurity, schooling, jobs, data centres and oversight, plus an AI voucher for SMEs and freelancers. The national AI plan wants over half of companies using the technology by 2030.
Swiss Federal Chancellery — Switzerland AI policy for federal offices follows a new Federal Chancellery two-year roadmap keeping six of eight planned measures after funding shrank to 2 million francs. Work centres on mapping process gains, an in-house generative AI system and legal groundwork for administrative use. A handbook and office marketplace were scrapped, with the current project database serving this government AI programme instead.
COP31 Presidency (Türkiye), Ministry of Environment, Urbanisation and Climate Change — Türkiye, hosting November's COP31 climate summit in Antalya, opened work at the UN General Assembly on the Antalya Pledge on Artificial Intelligence, a voluntary pact asking states to line up datacentre build-out with planetary warming targets. Türkiye AI policy here runs through the UN's International Telecommunication Union, co-writing a text due before the summit. President-Designate Murat Kurum argued that states should fix artificial intelligence energy rules and firms reveal how much power they consume.
Supreme Court of the Russian Federation — Supreme Court chairman Igor Krasnov signed the Concept for introducing artificial intelligence technologies into court proceedings on 6 September, advancing Russia AI regulation. Judges keep the last word. These judicial AI rules target up to 95% of the court system using AI by 2030, with machine-drafted texts forming 30% to 50% of procedural documents.
National Cyber Emergency Response Team (PKCERT) — Pakistan's National Cyber Emergency Response Team has shut officials out of feeding public AI tools any classified paper, work email, software code or citizen record, a Pakistan AI policy set in the National Cybersecurity Handbook 2026-27. Only departmentally approved tools may be touched, names stripped from prompts and files, and slips reported to departmental IT or cybersecurity teams. The public sector AI restrictions also bar handing passwords, administrative logins or API keys, and unapproved AI add-ons.
Malawi Communications Regulatory Authority — Fabricating content with artificial intelligence tools now risks prosecution in Malawi under a September 2026 caution from the Malawi Communications Regulatory Authority (MACRA). Officials cited Section 60 of the Penal Code, which bans publishing statements capable of provoking public alarm, and framed this AI misinformation warning around any digitally manipulated material. MACRA is tracing those behind such posts, noting exposure to criminal and civil penalties for AI-generated disinformation.
Ministry of Digital Transformation of the Republic of North Macedonia — A first draft national AI strategy from North Macedonia's Ministry of Digital Transformation is open for comment via the ENER portal, ahead of planned public hearings. Stefan Andonovski anticipates sign-off before 2026 closes, with the blueprint paving the way for a later artificial intelligence use law stressing responsible, values-driven deployment over restrictions on building the technology itself.
Office of the New York State Attorney General — New York Attorney General Letitia James has told staff inside artificial intelligence companies to route AI whistleblower complaints — anonymously if preferred — through the Office of the Attorney General's secure portal. The alert leans on the Responsible AI Safety and Education Act, in force from January 1, 2027, the SHIELD Act and the federal Computer Fraud and Abuse Act, the scaffolding of this AI policy. James said technology's promise excuses nobody from laws protecting New Yorkers.
Asamblea Nacional del Ecuador — Ecuador AI regulation has stalled: the Asamblea Nacional voted 88 in plenary to archive the Ley Orgánica de Regulación y Promoción de la Inteligencia Artificial. Its text would have created duties around specialist hiring, technical infrastructure, data systems, staff training and supervisory machinery. Members shelving the AI law blamed implementation difficulty and clashes with statutes already in force.
Office of the Governor of Missouri — Missouri Governor Mike Kehoe has signed Executive Order 26-18, a Missouri AI regulation for automated licence plate reader systems at state bodies and state-funded police. Most plate and vehicle data must be erased permanently within 30 days; vendors may not sell, share or commercialise Missouri-generated data; and the AI policy bars artificial intelligence facial recognition from ALPR systems. Kehoe cited constitutional and privacy concerns over indefinite storage of law-abiding citizens' movements.
Communications Regulatory Authority (Iran) — Iran's Communications Regulatory Authority now grants artificial intelligence permits through one electronic counter at cra.gov.ir, the most concrete piece of Iran AI regulation yet. Firms lodge a request, then track it to preliminary approval and issue. Under the artificial intelligence licensing regime a holder may stand up compute and hosting, then market text, vision, speech and model-building products.
Hong Kong SAR Government — Hong Kong's Government will create a Commissioner for AI post within the Digital Policy Office, the 2026 Policy Address says, turning Hong Kong AI policy support for bureaux on risk governance into a named job. The Chief Secretary for Administration will coordinate the new AI oversight role.
Ministry of Information, Communications and Technology (Eswatini) — Eswatini's Ministry of ICT has hired a consultant to draft the nation's first National AI Strategy. Minister Savannah Maziya announced this at a regional validation workshop in Ezulwini, calling it timely for building a framework guiding AI adoption and development, alongside a parallel COMESA-led regional initiative.
Federal Government of Belgium (Minister Vanessa Matz) — Vanessa Matz, the federal modernisation minister, has handed a draft national AI strategy to cabinet colleagues and the federated entities. Belgium AI policy must now win over coalition partners and regions before this proposed national framework gets adopted.
Italian Republic — Decreto Legislativo 160, dated 9 settembre 2026, squares national law with Regulation (EU) 2024/1689, Europe's artificial intelligence rulebook, and marks the sharp end of Italy AI regulation. It prescribes how police may run AI systems and pins down who bears civil and criminal liability. Gazzetta Ufficiale, Serie Generale 214 carried the Italian AI law on 15 September 2026, biting from 30 September 2026, after the Council of Ministers cleared both adaptation decrees on 4 August 2026 under article 24 of Law 132/2025.
Senado de la República Dominicana — Dominican Republic AI regulation now moves forward as one bill after the Senate Commission on Higher Education, Science and Technology folded two rival drafts together, citing heavy overlap. Commissioners combined the Ethical and Safe Use of Artificial Intelligence bill with a separate AI policy guidelines proposal into one framework spanning design, deployment and oversight across all sectors. The goal: a single, efficient consolidated AI legislation text, not two competing measures.
Board of Grievances — Saudi Arabia's Board of Grievances has adopted an AI governance framework judicial administration bodies will follow, signed by president Dr Ali bin Ahmed Al-Ohaydib. It covers artificial intelligence across court processes, administration and online services: secure data handling, accountable use, alignment with domestic law. This AI regulation in the judiciary builds on principles published in March 2026, with checks assigned to the Office of Artificial Intelligence and Emerging Technologies. Approval coincided with the UNESCO Global Forum on the Ethics of AI in Riyadh.
National Cybersecurity Standardisation Technical Committee (TC260) — The National Cybersecurity Standardisation Technical Committee issued the AI Safety Governance Framework 3.0 as the 2026 National Cybersecurity Awareness Week opened. In this China AI regulation update, risk categories are re-sorted and safeguards refreshed within the classify-respond-govern structure. AI safety rulebook versions 1.0 and 2.0 date from 2024 and 2025.
Islamic Consultative Assembly (Majlis), Iran — Iran's parliament speaker, Mohammad Baqer Qalibaf, has sent the National AI Development Law to the government, and Iran AI regulation now bands systems as low, limited, high or very high risk. Which of the four bands a system lands in turns on transparency, privacy, employment and public health. A National AI Organisation answering straight to the president will carry the AI risk rules, and a security assessment centre must publish results every six months.
Government of Côte d'Ivoire — Minister Djibril Ouattara presented the Plan d'accélération 2026–2028 as Impact IA's second edition got under way in Abidjan, placing national AI policy centre stage. The 40 priority schemes hang on seven pillars, one pledging a home-built artificial intelligence for citizens' wellbeing. Broadband, e-government, innovation, cybersecurity, digital skills and postal e-commerce fill out this state AI programme.
Government of Peru (Law 31814 regulation) — Law 31814's regulation reached its first compliance deadline, turning Peru AI regulation from promise into enforceable obligation. Anyone developing or deploying AI systems in health, education, justice, security, or economy and finance now answers for them legally under the AI law.
Superintendencia Nacional de los Registros Públicos (Sunarp) — Peru's national public registry authority, Sunarp, has approved an AI policy for public registries by resolution dated 4 September, holding every AI tool to human supervision, accountability, transparency, reliability, security, data protection and non-discrimination. Public registrars keep their judgement, responsibility and answerability for registry decisions, and Sunarp already runs JULIO, an in-house AI platform with modules for registry resolutions and binding precedents.
Digital Ministry of Malaysia — Malaysia's Digital Ministry has finalised the draft of the country's Artificial Intelligence (AI) Governance Bill, advancing Malaysia AI regulation toward Cabinet submission ahead of an expected Parliament tabling by the first quarter of 2027. The legislation proposes a Central AI Authority tasked with overseeing compliance and enforcement, forming one strand of a wider framework that blends technical standards, rules and statutes. The Bill represents the legal component of that broader structure, built to combine binding law with softer standards-based governance.
Instituto Costarricense sobre Drogas (ICD) — Costa Rica's Institute on Drugs (ICD) imposed a total, immediate AI ban on processing its own data, inputs or information, through circular CIRCULAR-ICD-DG-016-2026 issued on 28 August by director general Jorge Torres. It revokes an April 2025 directive permitting regulated use, as Torres cites a social and national-security duty to guard its information.
National Information and Communications Technology Authority — Papua New Guinea's National Information and Communications Technology Authority wants three fresh Bills widening control over data protection, digital identity and AI, tied to a broader Papua New Guinea AI regulation push. A proposed risk-based model targets banned uses and high-risk tools with bias testing, audits and disclosure duties, plus mandatory human review touching rights or public services. Stakeholder input and legislative passage remain ahead for this AI governance framework.
Procuración General de la Nación (Ministerio Público Fiscal) — Argentina's Attorney General signed Resolution PGN 64/26 on 4 September, setting AI rules for prosecutors and staff across the national Public Prosecutor's Office. Reserved information, live-case material and personal or sensitive data stay out of public AI tools such as ChatGPT until individual use is regulated, while permanent practical AI training and secure in-house tools are ordered. An internal survey beforehand found 90.4% of staff wanted a full training plan.
Supreme People's Court — The Supreme People's Court issued judicial guidelines for AI-related disputes, holding that recognisable digital replicas of a person — cloned faces and cloned voices included — may not be made or distributed without that person's consent. The guidance also reaches algorithmic price discrimination and AI-generated false information, and leaves service providers liable where they fail to act promptly after notice that their systems produced infringing content. Someone whose face has been altered and used to spread false and defamatory sexual claims now has a route to court; the court's vice-president, Tao Kaiyuan, cast this China AI regulation as seeking to balance development and security, with room left to clarify questions on which consensus is still thin.
Supreme People's Court — Beijing's top court has moved to govern AI disputes adjudication with new rules from the Supreme People's Court: the Opinions on Lawfully Adjudicating Cases Involving AI Disputes (Fa Fa [2026] No. 10), China's first nationwide judicial guidance on such matters. Its 24 articles set tort standards for deepfake harms like face-swapping and voice cloning, define intellectual property protections for machine-generated works, and fix evidentiary benchmarks. Criminal penalties are also spelled out for AI-enabled offences, including fraud and defamation, closing gaps in AI-related litigation.
Immigration and Refugee Board of Canada — The Immigration and Refugee Board of Canada now applies a practice notice on AI in refugee hearings across all four divisions. No party may have AI create or materially alter personal evidence, whether witness statements, affidavits or basis of claim narratives. Under these tribunal AI rules, filers answer for submissions, must verify AI-generated content and must disclose AI transcription or translation.
Quebec courts — The Quebec Court of Appeal, Superior Court, Court of Quebec and municipal courts jointly set generative AI rules for judges. Substantive legal reasoning, characterising or analysing facts, weighing evidence or credibility, deciding outcomes and drafting those passages stay off-limits; correcting, revising, rephrasing or translating a judge's own text, some administrative work and searching specified sources are allowed. These judicial AI guidelines followed a March report of a Quebec ruling citing nonexistent case law.
Government of Djibouti — Officials outlined directions for the first National Artificial Intelligence Strategy (SNIA) for 2026–2030, the core of Djibouti AI policy. Six pillars shape it: governing data and AI, cybersecurity, skills, research, innovation and inclusion. Only a sketch of this AI roadmap is public.
Intergovernmental Authority on Development (IGAD) — The Intergovernmental Authority on Development (IGAD) held two-day talks in Addis Ababa, Ethiopia, towards a Regional Artificial Intelligence Framework and Strategy and shared regional AI policy. Executive Secretary Workneh Gebeyehu opened discussions meant to yield a draft steering Member States on AI for sustainable growth. The Food Systems Resilience Programme for Eastern and Southern Africa underwrites the cross-border AI governance.
Senate of Paraguay — A draft law before Paraguay's Senate from Senator Lizarella Valiente introduces AI-generated content identification for synthetic images, voices and videos altered using artificial intelligence. It defines 'digital imitation', binds platforms to new duties, and adds synthetic media detection safeguards against unauthorised use of someone's likeness. Debate is pending.
Government of Sweden — Sweden AI regulation gained an inquiry under Dir. 2026:116: special investigator Lisa Englund Krafft will settle how the public access principle applies when agencies run artificial intelligence and external cloud services. She must map practice, decide if material in such tools counts as official records, and test whether rules for public-sector AI obstruct uptake, reporting by 15 November 2027.
Ministère délégué chargé de l'Économie numérique et de l'Innovation (MDENI) — The Ministry Delegate for Digital Economy and Innovation (MDENI) unveiled the first National Artificial Intelligence Strategy (SNIA) 2026-2030, which sets the direction of Djibouti AI policy, on day two of the first National AI Forum, with UNESCO and ESCWA backing. The national AI policy roadmap prioritises inclusion, innovation, research, training, cybersecurity and governance.
Government of Hungary — Hungary AI regulation now falls under the Minister of Science and Technology rather than the Minister of Economy and Energy, after a new government decree strips market surveillance duties from the latter and hands them to the former. Government Decree No. 139/2026 (VIII.31.) makes the switch official, amending the earlier Gov. Decree 90/2026 (V.13.) through a freshly inserted Section 168/A. The reassignment had effectively already happened before this paperwork caught up, since the supervisory body's own website was already presenting itself as answering to the Ministry of Science and Technology. The decree simply confirms, on paper, which ministry now carries responsibility for policing artificial intelligence products in the market.
Tribunal Superior Eleitoral — Brazil's superior electoral court voted 5-2 to define a political-ad deepfake as synthetic content made or altered by AI or similar technology, that convincingly creates, copies or changes a living, dead or fictional person's image, voice or speech. These political deepfake rules govern campaign material before the 4 October elections, including the roughly 1,000 political advertisements broadcasters must carry daily. The same ruling spared Senator Flávio Bolsonaro penalties over AI-generated images of his imprisoned father, because the party convention that showed them counted as a private event, outside public advertising.
State Duma — Russia's first federal AI law took effect on 1 September 2026, after the State Duma passed it on 8 July. It governs AI development, deployment and use, and separates 'sovereign' models, made entirely by a Russian company and developed, trained and stored in domestic data centres, from 'national' ones that may use foreign open-source components. Both gain status only once the government confirms they fit what the law calls traditional Russian spiritual and moral values.
Český telekomunikační úřad — Czech AI regulation now has a dedicated enforcer: Český telekomunikační úřad, the country's telecoms watchdog, has opened a standalone unit to police artificial intelligence systems. Anna Hroudová heads the team, having moved over from the Confederation of Industry and Transport of the Czech Republic, joined by three colleagues covering law, technology and analytics. Once enabling legislation clears parliament, the unit inherits authority to enforce the European AI Act, the bloc's framework for EU AI Act implementation, giving the regulator real supervisory bite over AI in practice.
Digital Agency — The Digital Agency's guideline on procuring and using generative AI in public administration took effect in version 2.0 on 1 September. This Japan AI regulation now covers voice as well as text input and image or audio output. Ministries must appoint a Chief AI Officer, and autonomous agents need reporting to the CAIO before launch under the public-sector AI rules.
Asamblea Nacional del Ecuador — Assemblywoman Cecilia Baltazar tabled the Proyecto de Ley Orgánica para el Desarrollo, Uso Ético y Gobernanza de la Inteligencia Artificial before the National Assembly. The Ecuador AI regulation proposal sets risk tiers, governance structures, innovation incentives and sanctions when AI systems misuse personal data, and amends the Código Orgánico Integral Penal. This AI legislation is at an early stage; the Assembly archived an earlier AI bill on 15 September 2026.
Red de Integridad y Estado Abierto / ECLAC — Chile's Red de Integridad y Estado Abierto and the UN Economic Commission for Latin America and the Caribbean are forming a network for independent public-sector AI oversight. Public institutions will first register the AI systems they run and self-assess against ECLAC's standard, before verification and oversight; the Comptroller General alone keeps audit powers, and algorithmic transparency leads the agenda.
US National Archives and Records Administration — The US National Archives and Records Administration's August memo (AC-11-2026) tells agencies that using AI does not, 'in and of itself', produce federal records. Whether AI material must be kept turns on factors like an agency relying on it for decisions, using it for official business, circulating it or building it into an agency system; NARA recommends formal AI policies.
German Federal Government — Berlin now hosts a freshly launched AI safety institute: AISI Germany, folding BSI's cybersecurity expertise together with the Federal Network Agency's former remit over operational safety of AI systems, all under one national AI safety body. A June 2026 resolution from the National Security Council brought it into being, with joint command shared between the Federal Ministry of Digital and the Federal Ministry of the Interior. Holding no sanctioning power, it instead weighs each AI model's dangers against its upsides, steering government, administration and industry as an analytical hub.
Gobierno de la Provincia del Chaco — Governor Leandro Zdero of Argentina's Chaco province signed Law 4250-J, an AI deepfake law, on 26 August, putting manipulated or synthetic digital content and bogus social-media profiles into the misdemeanours code. Doing so without informed consent risks up to 60 days' arrest or fines up to 15 minimum salaries, 90 days or 18 if shared; plainly artistic work, parody and public-interest criticism that cannot pass as real are exempt.
Prime Minister of Vietnam (Thủ tướng Chính phủ) — Vietnam's national AI strategy was approved on 28 August 2026, when the Prime Minister signed Decision 1671/QĐ-TTg approving the National Strategy on Artificial Intelligence to 2030 with a vision to 2045. Targets include a top-three ASEAN standing in AI R&D by 2030, a top-10 rank across Asia by 2045, and 100% of online public services integrated with AI. The plan drives AI transformation across governance—problem identification, design, organisation, operation—while building at least five shared national AI platforms and eight Vietnamese-language AI models.
Presidente da República / Instituto de Modernização Administrativa (IMA) — Presidential approval has cleared a $6 million outlay for an AI platform for public administration in Angola, formalised through Despacho Presidencial n.º 342/26 under President of the Republic João Lourenço. The order launches a competitive bidding process alongside a governance framework, formal policy and countrywide programme to prepare civil servants for embedding AI into government services. Once live, the system will handle virtual assistants and automate routine administrative workflows.
Government of the Slovak Republic — Slovakia has approved a draft act placing state administration of artificial intelligence under the Ministry of Investments, Regional Development and Informatisation (MIRRI), naming it the single contact point and central market surveillance authority for AI systems. This Slovakia AI regulation splits sector-specific duties among the National Security Authority, the Data Protection Office and the Justice Ministry, and requires MIRRI to build a regulatory sandbox for developing and testing AI under the new AI oversight law. The law takes effect 1 January 2027.
Islamic Consultative Assembly of Iran and the Guardian Council — Iran's Islamic Consultative Assembly voted through amendments to the National AI Development Plan at a session on 28 Mordad 1405. This Iran AI policy, formally the National AI Development Plan, then went to the Guardian Council, which examined the amended text on 4 Shahrivar 1405 and declared it consistent with Islamic law and the Constitution, clearing its last legal hurdle.
Government of Mongolia — Mongolia AI regulation advances as the Cabinet sends a draft Data Law to Parliament, proposing one legal structure for data creation, collection, processing, exchange, storage, infrastructure and security. Deputy Prime Minister T. Dorjkhand and Digital Development Minister Ch. Nomin presented the bill on 24 August 2026, citing global AI policy framework shifts and the need to build a regional data hub.
Ministry of Justice — Clearer limits now govern AI legal services in Japan, as the Ministry of Justice broadens earlier guidance beyond contract disputes to cover AI-powered legal business support more widely. A provider sidesteps unauthorised practice under Article 72 of the Attorneys Act if its tool stays neutral toward contested cases with adequate governance safeguards; the 2023 edition addressed only agreement-related matters.
Presidency of the Republic of Türkiye — The Artificial Intelligence Action Plan for 2026-2030 took effect through a presidential circular signed by Recep Tayyip Erdoğan and published in the Official Gazette. Now binding Türkiye AI policy, it requires public institutions to spend at least 2 per cent of investment budgets on AI projects and promises AI literacy training for five million citizens across 81 provinces within two years. A national programme will train 10,000 specialists and 100,000 application professionals under the AI plan.
Republic of Cyprus — Cyprus AI regulation reaches draft stage: the Artificial Intelligence Law of 2026, open for comment until 16 September 2026, carries the European Union's AI Act into domestic law. An Artificial Intelligence Regulatory Directorate would sit inside the Office of the Commissioner for Communications, with named competent authorities, oversight of AI systems and administrative fines for breaches. The national AI rules allow a sandbox for responsible innovation.
Government of Guinea / UNDP — Hiring is underway for Guinea AI strategy work, backed by the United Nations Development Programme. TREMPLIN, UNDP's local project, wants a consultant to write the National Artificial Intelligence and Data Strategy (SNIAD 2026-2031), building on the AILA diagnostic method and a 10-year roadmap agreed in December 2025. Applications closed 31 August 2026, starting the process that shapes how Guinea manages AI and data policy.
Presidency of Türkiye (Cumhurbaşkanlığı) / Ministry of Industry and Technology — Türkiye AI policy for 2026-2030 launches under the Türkiye Artificial Intelligence Action Plan, adopted via Presidential Circular No. 2026/9 in Official Gazette No. 33344. Four guiding principles—Recognise, Benefit, Produce and Govern—shape goals to grow 10000 advanced specialists and 100000 wider AI professionals, push public investment toward AI projects at a 2 percent minimum, and pull USD 10 billion from private-sector backers. A National AI Board plus a National AI Ethics Board join sandboxes spanning five priority sectors, with the Ministry of Industry and Technology steering this Turkish artificial intelligence strategy overall.
Ministry of Information, Communications, Science, Technology and Innovation (MICSTI) — Lesotho now has a finalised national AI policy, giving the country rules for building, rolling out and using artificial intelligence responsibly. Under Lesotho AI policy, priorities include boosting the economy, sharpening public-service delivery, driving innovation, expanding infrastructure, building skills, integrating AI across sectors, and protecting ethics and human rights. A three-tier national AI framework backs this up: a policymaker, an independent regulator, and a Data and AI Committee.
Government of Punjab — Chief Minister Maryam Nawaz has unveiled Punjab AI Vision 2029, an AI policy Pakistan hopes will make the province South Asia's top performer within three years. Six pillars anchor the scheme: a pioneering AI Office and a Delivery Unit officials call the world's first driven by AI-enabled governance. Coverage spans Pakistan's largest province by population, steering artificial intelligence into administration, services and economic growth.
Government of Tunisia - Ministry of Communication Technologies — Sofiane Hemissi, Tunisia's Communication Technologies Minister, revealed draft Tunisia AI strategy priorities running 2026–2030, touching farming, health, energy, transport, mobility and public administration. Goals include bolstering sovereignty, safeguarding constitutional rights and hardening cyberspace against mounting threats. Lacking an Official Journal decree, this national AI roadmap stays provisional rather than enacted.
Government of Serbia — Serbia's Ministry of Science, Technological Development and Innovation aims to pass the country's first Serbia AI law by December 2026. Modelled on the EU AI Act, the draft sorts systems by risk into prohibited, high-risk, medium-risk and minimal categories, backed by a new AI Agency and a registry for the riskiest tools. The resulting Serbian AI regulation would give the country its own dedicated framework for governing AI use.
Consejo de Gabinete / SENACYT — Panama's Cabinet Council has approved the National AI Strategy through Resolution 89-26, drafted under SENACYT, setting objectives, programmes and priorities for ethical, responsible, inclusive and secure AI and aiming to make the country a trusted AI innovation hub in Latin America. The resolution works as a policy roadmap: it adds no testing, certification or product-authorisation requirements, and ASEP equipment homologation stays as it is.
Cabinet of Mongolia — A new National Council on Artificial Intelligence will drive Mongolia AI policy, headed by the Prime Minister. Cabinet chiefs from Digital Development, Education, Finance and Foreign Affairs join industry and professional-body delegates, backing wider Mongolia AI governance. Its brief: unify strategy and push AI adoption across public services, education and healthcare.
Prosecutor General's Office — Eight AI-powered functions will join the Maldives' Mafini case platform under PGO 2.0, an AI prosecutorial services plan from President Dr Mohamed Muizzu for the Prosecutor General's Office. Additions include a legal document analyser, identity analyser, and charge-drafting and summarising tools, hosted on sovereign infrastructure under the office's AI governance rules, aiming for paperless AI in prosecution by 2028.
Prime Minister of Vietnam (Thủ tướng Chính phủ) — Vietnam's Prime Minister issued Decision 1528/QĐ-TTg, launching the National Programme for AI Human Resource Development to 2030, orientation 2035. Mandatory AI skills training for cadres, civil servants and public employees starts at 15% coverage in 2026, rising in stages to a 90% floor by 2030. Curriculum design and cross-agency rollout across ministries and provinces sits with the Ministry of Science and Technology, lead body for AI workforce development.
Ministry of Justice and Ministry of Digital Transformation and Telecommunications — Sudan's Ministry of Justice and the Ministry of Digital Transformation and Telecommunications are preparing the Draft Law on Regulating Data and Artificial Intelligence for 2026, part of broader Sudan AI regulation to match fast technological change. Undersecretary Mawlana Ali Khidr revealed the AI governance legislation at a workshop on workplace AI use, calling it the legal basis needed to oversee AI adoption across government and business.
Internal Revenue Service — Internal Revenue Service governance for AI now runs through a revised IRM 10.24.1, the IRS Policy for Artificial Intelligence Governance, driving IRS-wide AI governance in tax administration toward faster, trustworthy builds and rollouts. The Chief Data and Analytics Officer, now Responsible AI Official, must apply stricter rules to high-impact systems and maintain an AI use case inventory. The framework aligns with OMB M-25-21 and other federal mandates.
Data Protection Authority (DPA) — Sri Lanka now requires public sector data compliance before the Personal Data Protection Act's core provisions take effect on 1 January 2027. Circular No. 01/2026 tells ministries, departments, councils and state enterprises to name data protection officers, run protection programmes and audit personal data, replacing a 2024 version under the country's personal data protection regulation.
Urząd Ochrony Danych Osobowych (Personal Data Protection Office) — Poland's Personal Data Protection Office (UODO) issued four checklists guiding Polish organisations through AI compliance checks under GDPR before building or deploying AI systems. Lists suit smaller firms, government bodies and others, plus a longer version covering EU AI Act duties like risk tiering and rights-impact reviews, supporting GDPR compliance for AI systems.
Ministry of Information and Communications Technology — Malawi's Ministry of Information and Communications Technology is pushing an AI Bill covering artificial intelligence use in government and private firms, part of a fresh Malawi AI policy. Minister Shadric Namalomba disclosed the move at a Heads of ICT meeting in Lilongwe, linking it to an upcoming National AI Strategy, with lawmakers due to see it before December 2026. Officials frame the regulatory framework for AI as balancing innovation, privacy and accountability.
Ministry of Digital Economy — The Ministry of Digital Economy brought the Democratic Republic of Congo its first DR Congo AI policy on 4 August 2026: officials, donors and specialists convened at Kinshasa cleared twin texts charting course to 2030, the Politique sectorielle de l'economie numerique alongside the Strategie nationale de l'intelligence artificielle. Minister Augustin Kibassa Maliba unveiled them. The national AI roadmap leans on competence, research, mining data and homegrown builds, reaching agriculture, health, education, security, transport, natural resource management and public services. Ratification awaits the Council of Ministers.
Ministry of Information, Communications and the Digital Economy — Kenya AI regulation takes a new shape in the draft Artificial Intelligence and Emerging Technologies Policy, 2026, now open for comment and grounded in the National AI Strategy 2025-2030. Foreign AI providers face extraterritorial reach, liability runs across the full AI value chain, and Sovereignty and Strategic Autonomy anchors this AI governance framework, mandating local storage of sensitive datasets. The draft carries no binding force; it signals direction only.
Consumer Protection and Technical Regulatory Authority (TTJA) — Under Estonia AI regulation, the Consumer Protection and Technical Regulatory Authority (TTJA) will serve as competent authority overseeing AI Act compliance. Its explainer covers in-scope goods, sectors and eight prohibited AI practices. TTJA may demand providers' training datasets, logs and source code under this national AI rulebook.
Government of Slovenia / AKOS — AKOS now enforces AI transparency rules in Slovenia under Article 50 of the EU Artificial Intelligence Act. Providers and deployers of AI systems that interact with people, or generate synthetic content, must flag the AI's involvement or the output's artificial origin. AKOS, Slovenia's Communications Networks and Services Agency, is also the country's AI Act disclosure obligations contact point.
Legislative Assembly of Puerto Rico — Puerto Rico's Ley 163-2026 gives AI image rights explicit footing by folding deepfakes and cloned voices into the Right to One's Own Image Law. Jenniffer González Colón's amendment defines 'image' to catch digitally altered audiovisual, auditory or multimedia output mimicking likeness, voice or identifiable traits, extending deepfake likeness protection beyond older standards. Existing fines and damages still apply.
Czech Telecommunication Office — From 2 August 2026 the Czech Telecommunication Office, among the market surveillance bodies the Czech Republic has designated, spelled out how the AI transparency rules of the AI Act bite. Anyone supplying a chatbot must tell people they are dealing with a machine, and supplier output carries a machine-readable mark; operators label only deepfakes and AI-made material touching the public interest. Biometric and emotion-recognition kit obliges its operator to notify every person subjected to it, and enforcement of Article 50 can fall to the European AI Office instead.
Presidencia de la República del Ecuador / SERCOP — Ecuador's president Daniel Noboa signed Executive Decree 461, reforming procurement rules to phase in AI in public procurement: automated early-warning systems, data mining, natural-language processing and machine learning. Procurement agency SERCOP will use them to spot tendering risks, irregularities and breaches, funded by a new special levy.
Government of Malta — Malta's new Artificial Intelligence Governance Council, led by Chief AI Officer Jonathan Gerada, now steers government-wide AI adoption in the public sector. The council unites public administration representatives to turn oversight decisions into concrete action, prioritising public value over cost reduction. This reflects Malta's approach to AI governance in government, favouring coordinated implementation over savings-driven deployment.
Ministry of Justice of the Republic of Serbia — Serbia's Ministry of Justice has drafted a Law on Personal Data Protection setting first-time AI data protection law rules for handling personal information within artificial intelligence systems, part of the country's broader AI governance regulation push. High-risk systems require an impact assessment plus the Commissioner's opinion before use; minimal-risk tools face no such demand. Synthetic audio and video need clear labelling, with public comment open until 10 September.
German Federal Government — Germany AI regulation now runs through KI-MIG, the AI Market Surveillance and Innovation Promotion Act, live since 29 July 2026 after Bundestag passage on 11 June and Bundesrat clearance on 10 July. BNetzA leads as watchdog, EU AI Office contact, and complaints handler under the EU AI Act. A built-in sandbox backs AI Act enforcement Germany, letting firms trial systems under supervision.
Ministry of Digital, Malaysia — Prime Minister Anwar Ibrahim has created AI Malaysia Berhad under the Ministry of Digital, making the National AI Office set up in December 2024 permanent with an implementation mandate. A Malaysia AI Safety Institute also launched to support national AI governance on standards and compliance, run AI safety evaluations and broker talent and technology partnerships.
Ministry of Digital — Malaysia AI policy takes shape under a new five-year strategy, the National AI Action Plan 2026-2030, branded AI Nation 2030, aiming to turn the country into a regional digital technology hub by 2030. It groups 28 initiatives into two tiers — 14 aimed directly at socioeconomic growth and 14 backing broader impact sectors — anchored on skilled talent, market-driven innovation, computing and data infrastructure, sound governance, and sustainable financing. The AI Nation 2030 roadmap launched alongside a new entity, AI Malaysia Berhad.
Polish Parliament and President of Poland — Poland's Sejm passed the Act on Artificial Intelligence Systems on 11 June 2026, the country's implementation of Poland AI regulation under the EU AI Act; President Andrzej Duda signed it into Polish law on 27 July 2026. It sets up the Commission for Development and Safety of Artificial Intelligence (KRiBSI) as market surveillance authority, creates regulatory sandboxes, and mandates AI-generated content labelling. The AI Systems Act Poland rules apply in stages from July 2026.
Ministry of Science, Technology and Innovation, Nepal — A single umbrella law taking shape at Nepal's Ministry of Science, Technology and Innovation would gather AI regulation in Nepal under one roof with robotics, cybersecurity, intellectual property and startups. Officials cleared the concept paper for the Science, Technology and Innovation Bill, intended to swap out patchwork existing rules for unified artificial intelligence governance across these linked sectors.
Ministry of Digital Development and Communications, Myanmar — Myanmar has stood up a new body to push AI technology development: the National Artificial Intelligence Development Department, housed within the Ministry of Digital Development and Communications. Parliament heard from the Deputy Minister that the unit is driving national AI capacity building via fresh initiatives. The country separately joined the World Artificial Intelligence Cooperation Organization, formalised at Shanghai's 2026 World AI Conference.
Legislative Assembly of Puerto Rico — Governor Jenniffer González Colón's Ley 140-2026 folds AI content in cybersecurity continuing education into Puerto Rico's Cybersecurity Law, an artificial intelligence coursework requirement for government workers. The mandate pairs AI instruction with privacy, system defence, data handling, cyberattack prevention and responsible adoption of new tools.
Cabinet of Ministers of the Republic of Azerbaijan — Prime Minister Ali Asadov approved priority sectors for AI development under Azerbaijan AI policy, each with an accountable state body. Ministries and agencies lead healthcare, education, agriculture, energy, transport, justice, ecology, finance (with the Central Bank), urban planning, public services (ASAN), and economy and industry. Each oversees adoption, talent and regulation in its field, making AI governance sector-led.
Government of Papua New Guinea — Papua New Guinea's parliament will debate amendments to the Cybercrime Code Act in November making AI deepfakes, synthetic voice impersonation and other malicious AI uses criminal offences. The bill also covers AI-generated child exploitation material, sexually explicit imagery, fraudulent cloned voices, digital identity theft and online scams. Acting ICT Minister Peter Tsiamalili said platforms and AI service providers would face liability for unlawful content.
Oireachtas (Irish Parliament) — Ireland's Oireachtas has handed the country a dedicated AI regulator: the Regulation of Artificial Intelligence Act 2026 (Number 31 of 2026) creates the AI Office of Ireland, Oifig IS na hÉireann, tasked with enforcing the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) locally. This Irish AI regulation assigns market-watchdog duties to named bodies, opens sandboxes for testing high-risk systems under real conditions, and rolls out a public register, one-stop contact point, complaint-handling steps and penalties, together forming the state's new AI oversight framework.
Spanish Data Protection Agency (AEPD) — The Spanish Data Protection Agency (AEPD) has issued Spain AI regulation guidance: a technical note on data accuracy, suitability and quality where Artificial Intelligence processes personal data. This AI regulatory guidance examines how GDPR accuracy and minimisation principles bear on AI systems, and recommends steps to secure data quality for AI training and operations.
Hellenic Parliament — Under Law 5321/2026, Greece rolls out Greece AI regulation via EU AI Act implementation for Regulation (EU) 2024/1689. Oversight splits between the Hellenic Data Protection Authority for market surveillance and the Hellenic Telecommunications and Post Commission as notifying authority, alongside a sandbox, an AI Observatory and a public-sector systems registry. Stripping transparency labels from deepfake content now carries criminal liability.
Consell General — The Consell General passed Penal Code revisions without dissent, tightening Andorra AI regulation: buying child sexual abuse imagery online, AI-generated included, draws prostitution-level penalties. Those AI content rules join a consent age lifted from 14 to 16. Justice and Interior Minister Ester Molné tabled the qualified laws.
Saudi Data and AI Authority (SDAIA) — The Saudi Data and AI Authority (SDAIA) has launched the National Framework for AI Risk Management in Riyadh, for government and private bodies. It sets a national method to identify, assess, treat and monitor AI-system risks, which SDAIA says differ from ordinary software because behaviour may shift in use and outputs resist explanation or reproduction. The framework is non-binding guidance.
Cabinet of Japan — Under Japan AI policy, the Cabinet adopted a second-phase AI Basic Plan on 14 July, replacing the December 2025 original, subtitled 'Japan AI Transformation: A Stronger Nation. Greater Prosperity.' It spans four pillars: faster deployment, stronger domestic development, governance leadership and an AI-powered society, pushing agentic, vertical and physical AI into healthcare, finance, education and disaster response. A new 'AI sovereignty' principle seeks less reliance on any single vendor or nation.
Government of Curaçao — Educators, scientists, firms, civic groups and officials are being invited to shape Curaçao's national AI strategy under a fresh push from the Government and its Regulatory Authority (RAC). Sithree van Heydoorn heads this effort, drawing on the UNESCO AI Readiness Assessment and Caribbean AI Policy Roadmap to steer the territory's AI policy development.
Autoriteit Persoonsgegevens (AP) — Two new instruments from the Dutch Data Protection Authority, AP, tackle AI GDPR compliance: technical guidance for developers and a self-assessment checklist letting organisations judge their own readiness before rolling out generative AI. Lawful bases, required protections and record-keeping duties under the GDPR run through both. AP separately revived its earlier consultation opinion on Microsoft 365 Copilot, tying personal data protection in generative AI back to that regulatory approach.
Prime Ministry (Jordan) — Jordan's government has barred ministries, public institutions, state universities, municipalities, Greater Amman Municipality and government-owned companies from using AI to produce or alter flags, seals, official photographs, sovereign symbols or national emblems in ways that affect authenticity or approved form. The Jordan AI regulation also bans publishing or adopting AI-compromised symbols or seals; government AI rules require the original approved versions.
Parliamentary Assembly of Bosnia and Herzegovina — The Parliamentary Assembly of Bosnia and Herzegovina has cleared ratification of the Council of Europe Framework Convention on Artificial Intelligence and Human Rights, Democracy and the Rule of Law, the earliest binding treaty anywhere on using artificial intelligence. Bosnia and Herzegovina AI regulation gains no home-grown statute: consent, carried in the Official Gazette - International Treaties number 5/26 of 9 July 2026, ties the state to an outside legal framework that institutions and later rules must deliver. Signature came on 9 December 2025 in Strasbourg, settling the terms of national AI oversight.
National Assembly of Nigeria — Nigeria AI regulation nears enactment via the National Digital Economy and E-Governance Bill, now at third reading. Lawmaker Shuaib Salisu says the draft grades AI systems by risk, tying deployment duties to each grade. Oversight would pass to the National Information Technology Development Agency, part of a wider AI governance framework taking hold nationally.
Department of Information and Communications Technology — AI deepfakes regulation is moving forward in Papua New Guinea, where officials have begun drafting rules against voice cloning, digital impersonation and other misuse of artificial intelligence. Planned changes to the Cybercrime Code Act would criminalise sexually explicit deepfake images, child exploitation material, scam calls using cloned voices, and related online fraud, effectively criminalising AI-generated impersonation. Parliament may take up the bill in its November 2026 sitting, though the text remains unfinished and no statute yet exists.
Judiciary of Trinidad and Tobago — Chief Justice Ronnie Boodoosingh has signed off on a Practice Direction on the Use of Generative Artificial Intelligence in Court Proceedings, giving Trinidad and Tobago's Judiciary a rulebook for AI use in courts. Staff may lean on the tool for drafting and admin only once outputs are checked by a person, fabricating evidence with it is banned outright, and its use must be flagged in any filing; reasoning through a case or writing a judgment stays strictly a human job for judicial officers, so generative AI in judicial proceedings stops there. Fabricated citations and confidentiality leaks were the risks driving the change.
Niyel / Governments of Mali, Burkina Faso, Senegal, Côte d'Ivoire, Guinea, Benin — Niyel, a West African advocacy group, has led six nations—Mali, Burkina Faso, Senegal, Côte d'Ivoire, Guinea and Benin—into a joint framework for Mali AI governance and regional artificial intelligence policy. It urges transparent systems, stronger data protection, homegrown innovation and digital sovereignty, meant to supplement rather than replace each country's own strategy, and went to the UN, the United Nations, as a written submission at its first Global AI Dialogue in Geneva.
Government of Burkina Faso — Six West African states—Burkina Faso, Senegal, Côte d'Ivoire, Mali, Guinea and Benin—now share a common Burkina Faso AI policy: Guidelines for Ethical, Inclusive and Responsible Artificial Intelligence, lodged with the UN Global Dialogue on AI Governance. It demands accountable, transparent, sovereign, inclusive systems reflecting local languages and African realities.
National Cybersecurity Authority (NCA) — New AI security rules for agents in Saudi Arabia bring agentic and generative technologies under review via draft AI Cybersecurity Guidelines from the National Cybersecurity Authority (NCA), open for comment until 5 August 2026. Four pillars structure the proposal: governance, defence, resilience, third-party risk — an early regional step on autonomous-system safeguards. No AI governance timeline has surfaced since.
Dutch government — Netherlands AI policy now includes an International AI Strategy, with the Dutch government setting out global cooperation on trustworthy AI. Three priorities anchor it: diplomacy on European and worldwide regulation, alliances safeguarding shared values, and stronger European AI infrastructure cutting reliance on outside powers, part of its Dutch AI strategy push.
Cabinet of Kenya — A new Standing Cabinet Committee on Artificial Intelligence, formed under William Ruto, will now steer Kenya AI policy. Its brief covers building safe-rollout governance frameworks, lifting productivity via AI adoption, and reinforcing public services, plus a broader Kenya AI regulation drive backing job creation and inclusive growth.
Direction interministérielle du numérique (DINUM), Direction interministérielle de la transformation publique (DITP), Direction générale de l'administration et de la fonction publique (DGAFP) — Five binding rules now govern AI usage for civil servants across French territory, under DINUM's guide with DITP and DGAFP on 26 June 2026 — France's first interministerial AI usage guide for public agents, Guadeloupe included. Staff keep accountability for AI-assisted decisions, route sensitive data only through approved systems, and disclose AI's substantial role in outputs. Copyright-respecting, ethical practice plus training complete the framework.
Ministry of Industry and Trade of the Czech Republic — A single point of contact for Czech AI regulation would sit with the Czech Telecommunications Office, tasked with leading market checks under a first-of-its-kind national statute now written by the Ministry of Industry and Trade to fold the EU AI Act into the law of the Czech Republic. Two other bodies, the Czech National Bank and the Office for Personal Data Protection, would supervise their own fields, and a sandbox for trialling artificial intelligence tools would open under the Czech Agency for Standardisation. Other ministries gave feedback in autumn 2025, and the bill went to cabinet on 26 June 2026; come September 2026 no minister had reviewed it yet, leaving this proposed AI Act Czechia measure stalled before reaching parliament.
Tallinn Administrative Court — Ten people challenging a felling permit were fined by Tallinn Administrative Court after filing a submission whose citations named Estonian scholars who do not exist and studies nobody could trace — the first penalty of its kind under emerging Estonia AI regulation. One paid 150 euros and nine paid 50 each, well inside the 3200 euro ceiling; Tallinn Circuit Court upheld the ruling and it became enforceable once the Supreme Court refused the appeal. Judges accept that parties may lawfully use such tools, yet hold them answerable for whatever they assert. Kaupo Kruusvee, who chairs the administrative court, would rather litigants wrote in plain words of their own, and while earlier warnings and reprimands over careless AI use in court are on record, no previous fine is known.
Ministry of Human Resources and Social Security, National Development and Reform Commission, Ministry of Industry and Information Technology, National Data Administration — China's four ministries have ordered nationwide rollout of AI in human resources and social security services under the Implementation Opinions on Accelerating AI Plus Human Resources and Social Security Application Development (Ren She Bu Fa [2026] No. 40), covering employment support, insurance administration, talent development, labour relations and HR functions. Officials in these fields must gain digital and AI-driven HR and social security services competencies, with staged benchmarks set for 2027 and 2030.
Falkland Islands Government — The Falkland Islands (Malvinas) Government has backed plans to outlaw AI-generated deepfake intimate images, endorsing an Executive Council paper that argued for a ban. Officials will next draft legislation spelling out offences for both producing and ordering such AI-generated non-consensual sexual imagery.
Direction générale de l'administration et de la fonction publique (DGAFP) — Talks on AI in the civil service began in France as the DGAFP met unions and employers across all three public-sector branches to negotiate a framework agreement on rolling out and using artificial intelligence in government work. The draft, tied to the 'Notre IA' plan, groups its provisions into six strands: guiding principles, ethics and transparency duties, oversight structures, dialogue with staff representatives, workforce training, and handling organisational change from AI deployment in public administration. Negotiators are due back at the table on 10 July 2026.
Ministry of Science, Technological Development and Innovation of Serbia — Serbia AI regulation is taking shape: a working group under the Ministry of Science, Technological Development and Innovation aims to finish the nation's inaugural artificial intelligence statute by December 2026. Systems face sorting into four bands—banned outright, high-risk, needing disclosure, or minimal concern—while officials plan a new AI Agency alongside a registry tracking high-risk tools. The European Union's AI Act supplies the template, tying into Serbia's own AI Development Strategy stretching to 2030.
Eesti.ai council, Government Office of Estonia — Estonia's Eesti.ai council, set up at Prime Minister Kristen Michal's initiative, agreed at its second meeting to create a digital identity for AI agents, which it calls an AI personal identification code. Michal backed letting agents act for a person, company or institution within defined, verifiable and auditable rights, such as only viewing data, preparing a document or paying up to a set ceiling, never all of a user's rights. Moving quickly, he said, would make Estonia the first country to grant AI agents an official digital identity.
Government of Panama (Executive Decree No. 11 of 2026) — Panama's Executive Decree No. 11 of 17 June 2026 regulates the national Amber Alert system, implementing Law 469 of 2025 with responsibilities, procedures, response times and agency coordination. It promotes AI in missing-child searches with mobile apps, CCTV, video analysis and geolocation, and creates an Early Activation Office for reports, case assessment and alerts.
Ministère de la Transition digitale, des Postes et des Communications électroniques — Burkina Faso's Ministry of Digital Transition, Posts and Electronic Communications, through SP-IVTEN, finalised a national AI roadmap covering 2026-2030. Six focus areas structure the plan: oversight, infrastructure, safeguarding information, skills-building, entrepreneurial research, and fair, inclusive deployment. The strategy links to RELANCE 2026-2030 and the 'AI for All' scheme, embedding AI policy in broader growth ambitions.
Government of Israel — 100,000 advanced AI accelerators and a freshly created National AI Directorate inside the Prime Minister's Office anchor the national AI programme Israel chartered under Government Resolution No. 4255, "Accelerating Artificial Intelligence in Israel and Establishing Global Leadership". Staff there subsequently expanded that charter into a broader AI strategy Israel now brands its National AI Action Plan, rolled out 13 August 2026.
National Cyber and Information Security Agency (NÚKIB) and Office of the Government — NÚKIB (National Cyber and Information Security Agency), with an inter-ministerial group under the Office of the Government, gives officials a Desatero — Ten Commandments — for using AI responsibly. Security, data protection, transparency and trust anchor its Guide for Ethical and Responsible Use of AI in Public Administration. Goal: safe, purposeful AI in government across Czechia's public sector.
Ministry of Digital Transformation of Ukraine — A new panel convened by Ukraine's Ministry of Digital Transformation held its inaugural session on 15 June 2026, tasked with producing AI regulation Ukraine can align to the EU AI Act by year's end 2026, ending the current reliance on voluntary self-regulation. Officials, business figures and civil society representatives jointly guide the resulting Ukrainian AI legislation.
UAE Cabinet — One new body, the Artificial Intelligence and Data Authority, now runs UAE AI governance after the Cabinet folded in three offices: the Office of Artificial Intelligence, Digital Economy and Remote Work Applications, TDRA's Digital Government Sector, and the UAE Data Office. Answering to the Cabinet, it sets strategy, rules, and compliance checks, and builds outside ties under national AI regulation.
Centre national d'investigations numériques — A Beninese TikToker posting as Fresnel was detained on 11 June 2026 over an AI-generated image pairing him with former head of state Patrice Talon. Article 576 of the Code du numérique criminalises malicious doctoring of photographs, audio or video, drawing five years' jail and a 25 million CFA franc fine. The Centre national d'investigations numériques had warned days before against exactly this synthetic likeness.
Legislative Assembly of Puerto Rico — Ley 105-2026, signed by Governor Jenniffer González Colón, a Puerto Rico AI regulation, obliges campaign ads, notices or messages fully or partly AI-made or AI-modified to show plain, visible labels. Election AI rules amend Campaign Finance Law and Electoral Code, treating AI as technology doing human-intelligence work, including generative systems for text, images, audio, video or other digital material.
Department of Information and Communications Technology — The Department of Information and Communications Technology and the Civil Service Commission have issued Joint Memorandum Circular No. 003, Series of 2026, principles-based public-sector AI rules for how Philippine agencies build, deploy and use AI. Agencies must match AI work to privacy, security, accountability, transparency and fairness and create oversight machinery, including an internal IT, legal and operational committee; it takes effect on publication.
Government of Rwanda — A National Artificial Intelligence Agency has cleared Cabinet in Rwanda, standing as the nation's first body built solely for AI regulation in Rwanda. President Paul Kagame chaired the 8 June 2026 meeting that formed it, charged with driving faster innovation, adoption, investment and oversight. Anchored in the 2023 National AI Policy, it will link efforts on AI governance spanning ministries, business, universities and outside partners.
Government of Niger — President Abdourahamane Tiani's Décret n° 2026-310/PRN/PM abolishes nine public bodies, among them the Haute Autorité de Protection des Données à caractère Personnel (HAPDP). HAPDP enforced Niger AI regulation, namely Loi 2022-59's automated-decision and AI rules. Its data protection duties pass to the Ministries of Justice and Interior.
Saudi Data and Artificial Intelligence Authority (SDAIA) — The Saudi Data and AI Authority (SDAIA) has issued 10 documents shaping Saudi Arabia AI regulation on ethical, responsible use. The set anchors national AI rules for deployment across the Kingdom, reinforcing SDAIA's governance role in the Year of Artificial Intelligence 2026. It covers deepfake guidelines, AI ethics principles, generative AI principles aimed at government entities and the public, a national data and AI professional standards framework, and an academic qualifications framework for the AI specialty.
Arcom (Autorité de régulation de la communication audiovisuelle et numérique) — France's media regulator Arcom now binds all 2027 presidential campaign material to strict AI disclosure duties under recommendation n°2026-05, covering AI-generated content in elections, published 7 June 2026. Labels reading "Contenu généré par IA" become compulsory wherever AI substantially shapes campaign material, while faking a declared candidate's likeness without their signed approval becomes illegal from 1 November 2026. Major platforms face a 1 October 2026 deadline to launch tools letting users flag suspect material, feeding a registry that refreshes weekly once the official campaign period begins, giving the public a running record of AI electoral content flagged during the race.
Innovation, Science and Economic Development Canada — In Toronto on 4 June 2026, Prime Minister Mark Carney rolled out AI for All, a national AI strategy setting Canada's five-year course on artificial intelligence. Officials expect CAD 200 billion added to the economy and 250,000 fresh AI-related roles, financed through incoming laws and programmes meant to keep rollout responsible. Public confidence, opportunity and Canadian sovereignty underpin the wider push toward AI adoption.
UAE Federal National Council — UAE's Federal National Council enacted the Digital Economy and Artificial Intelligence Content Regulation Law, effective 1 September 2026. This AI content regulation law mandates labels on AI-generated images, video and text on UAE-licensed platforms, plus registration and disclosure rules for data brokers. Penalties reach AED 5 million per breach under this AI-generated media regulation, the GCC's first such federal framework.
Ålands lagting — Ålands lagting has passed the Landskapslag om tillsyn över artificiell intelligens (AI Oversight Law), transposing the EU AI Act into legislation for the autonomous region. Following two readings, the statute names Datainspektionen as the supervisory authority responsible for enforcement, giving the body powers over AI system oversight and compliance across Åland's jurisdiction. The framework applies AI system compliance regulation modelled on the AI Act's structure, adapted to the islands' devolved legal system.
Supreme Court of India, Artificial Intelligence Committee — Machines will not decide case outcomes or score bail and recidivism risk, under draft rules on AI in courts from the Supreme Court of India. Justice P S Narasimha's Artificial Intelligence Committee drew up the 2026 Regulations for Use of Artificial Intelligence (AI) in Courts, requiring disclosure of machine-generated content in judicial proceedings and backing a standing Apex Body at the Supreme Court to govern such use. Stakeholder feedback closed on 20 June 2026.
Digital Brunei Council, Ministry of Transport and Infocommunications — Brunei Darussalam merged its AI policy for the public sector into Digital Brunei 2030: The Digital Brunei Transformation Plan, launched 2 June 2026 by the Digital Brunei Council, joining four digital strategies under one roof, including the cross-cutting Data and Artificial Intelligence Strategy. Priorities range across governance, infrastructure, workforce skills, innovation and global ties, seeking a firmer AI ecosystem inside government. Metrics track trained officials, AI-backed pilots and overall public-sector AI maturity.
Government of National Unity (Libya) — Libya's prime minister, Abdulhamid Dbeibah, oversaw adoption of a National Charter for AI Ethics in Tripoli on National Technology Day, together with a national AI strategy for 2026–2030. The government presents this AI ethics charter as a move to regulate AI technologies; the source does not detail its individual provisions.
Government of National Unity — A new Libya AI policy sets four numeric benchmarks for 2030: 80 percent of government entities running AI solutions, 10000 civil servants trained, 100 AI startups backed, and 50 percent of government transactions automated. Tripoli's Government of National Unity approved these goals through its National Strategy for Artificial Intelligence 2026-2030, paired with a companion National Charter for AI Ethics. Together the two documents anchor AI adoption in the government sector as the plan's defining ambition.
New Zealand Parliament — New Zealand's Social Security (Modernisation) Amendment Bill has passed its third reading in Parliament, widening automated decision-making in welfare. Social Development and Employment Minister Louise Upston says automated welfare decisions will cover straightforward cases needing no discretion; the Bill went through under urgency on Budget day.
National Agency for the Information Society (AKSHI) — Albania's National Agency for the Information Society has put a first draft of Albania AI regulation to public consultation, running 29 May to 26 June 2026. The proposed draft law on artificial intelligence runs to nine chapters, 106 articles and 13 annexes, closely following EU Regulation 2024/1689 by grading systems on risk, banning manipulative or social-scoring uses outright, and layering heavy technical and governance duties onto anything rated high-risk. AKSHI would coordinate enforcement, and the text bars any legally damaging decision resting solely on real-time biometric identification.
Centre national d'investigations numériques — Benin's Centre national d'investigations numériques has spelled out penalties for AI image manipulation: reshaping, inventing or circulating another's likeness without consent. Such AI-altered likenesses breach image rights and draw prosecution under article 576 of the Code du numérique, which permits five years' jail and a 25 million FCFA fine, about 44,400 dollars.
Presidency of the Bolivarian Republic of Venezuela — Decreed 29 May 2026 by Delcy Rodríguez, Venezuela's new Vice-Ministry of Artificial Intelligence and Productive Efficiency sits under the Ministry of Science and Technology, extending AI in government toward industrial modernisation. A parallel Technology and Data Management Centre applies predictive-maintenance and output-optimisation software to hydrocarbons. Officials frame this as groundwork for broader artificial intelligence in the public sector.
National Agency for Information Society (AKSHI) — The government opened public consultation on its draft Law on Artificial Intelligence, which bases Albania AI regulation largely on the EU AI Act. The draft sorts AI systems by risk, bans manipulative practices and social scoring, mandates human oversight of high-risk systems, makes AKSHI coordinating authority and adds a regulatory sandbox. The AI rules allow real-time biometric identification in public spaces only with a court order for serious crimes.
Ministry of Public Administration of Montenegro — Montenegro AI policy advanced on 27 May 2026 when a Working Group under the Ministry of Public Administration, aided by UNDP, held its fifth meeting to pore over the strategy's initial blueprint for 2026-2030. Participants debated ambitions at both strategic and operational levels, plus yardsticks for gauging success. Attention centred on building digital infrastructure, sharpening skills, and shaping regulatory frameworks — pillars of the emerging national AI framework.
Government of Spain, Ministry for Digital Transformation and Public Service — The Council of Ministers forwarded the draft Organic Law for the Good Use and Governance of Artificial Intelligence to the Congress of Deputies, aligning Spain AI regulation with the EU AI Regulation. Penalties climb to 35 million euros or 7% of global turnover; AESIA, AEPD and CGPJ oversee compliance; a national AI sandbox, a public-body AI catalogue and a public-sector AI delegate follow. By September 2026 this AI legislation remained before parliament, unpassed.
Joint Court of Justice of Aruba, Curaçao, Sint Maarten and Bonaire, Sint Eustatius and Saba — Fairness, independence and openness underpin an emerging AI policy for judges at Curaçao's Joint Court of Justice. A panel assembled during the annual Knowledge Week produced this judicial AI governance framework, spanning bench and personnel alike. Directors took up the draft for scrutiny in May 2026.
Council of Ministers of Algeria — Algeria's Council of Ministers, under Prime Minister Sifi Ghrieb, has approved a national AI strategy for 2025-2030 spanning six pillars: research, skills, infrastructure spending, ecosystem growth, oversight rules, and priority fields such as health, energy and farming, plus sovereign AI models. This Algeria AI strategy, an artificial intelligence roadmap for Algeria, draws on 74 master's programmes at 52 universities training 57,000 computer science students.
Government of Benin / Presidency of the Republic — President Romuald Wadagni has given Benin a dedicated AI ministry, the Ministry of Digital Transformation and Innovation (MTDI), tasked with the national AI strategy. Mahuna Akplogan, a technologist with a PhD in AI, becomes Benin's first minister holding formal AI policy mandate responsibility. The move underlines Benin's bid to lead AI governance across the region.
Gouvernement de la République Gabonaise / Conseil des Ministres — Gabon's Conseil des Ministres approved a decree creating the Direction Générale des Technologies Émergentes (DGTE), part of a wider Gabon AI regulation push under the Ministry of Economie Numérique. The DGTE will draft and oversee public policy on digitalisation, technological innovation, and responsible use of artificial intelligence across République Gabonaise, with central departments and decentralised offices in every province.
Government of the Republic of Slovenia — A newly formed AI ethics council, the National Council for Ethics in Artificial Intelligence, now has its full roster of members named by the Government of the Republic of Slovenia, acting under authority granted by the ZIUDHPUI law. This independent panel of specialists will judge how artificial intelligence affects society, democracy and ethics, offering guidance to shape responsible public-sector adoption. Beyond issuing formal advice, the body carries a mandate to weigh the legal, environmental and social consequences of AI tools in use, functioning effectively as an ethical AI oversight body for the state.
Executive Yuan — Premier Cho Jung-tai now heads Taiwan AI strategy efforts via the newly formed National AI Strategy Special Committee, steering AI-related affairs government-wide. First convened 23 June 2026, it anchors national AI governance within the Cabinet.
Ministry of Communication, Technology and Innovation (MoCTI) — Sierra Leone AI strategy plans advance under the Ministry of Communication, Technology and Innovation with the World Bank Group, building the nation's first such blueprint. A draft AI Readiness Diagnostic Report measures preparedness. Media, civil society, government and business spent three days shaping this national artificial intelligence policy around homegrown culture.
Isle of Man Government — Isle of Man AI regulation lets businesses own AI training data as lawful assets after Data Asset Foundations legislation gained Royal Assent, the world's first statutory framework governing data assets. That permits pricing, licensing and safe exchange with collaborators. Follow-on rules and a world-first statutory Data Asset Register are being drafted; Enterprise Minister Tim Johnston argues such AI rules put the island ahead in the worldwide data economy.
Ministry of Communications of Turkmenistan — Turkmenistan AI strategy groundwork is underway, led by the Ministry of Communications with United Nations Development Programme (UNDP) assistance toward a national artificial intelligence policy. A reviewed draft fixed priority directions under the State Program for the Development of the Digital Economy for 2026-2028. Farming, transport, schooling and medicine sit within scope, adoption still pending.
UAE Cabinet — UAE Cabinet backing now covers a federal framework pushing 50% of ministry operations into agentic AI government services within two years, part of the Agentic AI Project. Each entity builds a team, led by its minister, fixing targets and indicators for this agentic AI public sector implementation. Duties sit clearly with every federal body named in the plan.
Holy See — Pope Leo XIV has instituted an Interdicasterial Commission on Artificial Intelligence, the forum for Holy See AI policy. Cardinal Michael Czerny, prefect of the Dicastery for Promoting Integral Human Development, signed the rescript, released by the Holy See Press Office; papal assent came at a 3 May 2026 audience. Seven bodies take part: that one, those for the Doctrine of the Faith, Culture and Education and Communication, and the Pontifical Academy for Life, of Sciences and of Social Sciences. Whichever body coordinates — a renewable yearly turn, Czerny's first — must ease cooperation and exchange on artificial-intelligence work, Vatican AI governance rules for use inside the Holy See included.
Dicastery for Promoting Integral Human Development, Holy See — The Holy See has launched an Interdicasterial Commission on Artificial Intelligence (ICAI) for Vatican AI governance church-wide. Cardinal Michael Czerny, Dicastery for Promoting Integral Human Development Prefect, signed the founding decree after Pope Leo XIV's approval on 3 May 2026. Drawing on seven dicasteries and pontifical academies, the body will pool knowledge and set shared AI policy for the central administration.
Agencia Nacional de Inteligencia Artificial (ANIA) — El Salvador AI policy now points toward sovereign tech independence, with the Agencia Nacional de Inteligencia Artificial (ANIA) projecting AI could drive nearly 20% of GDP by 2031. The Estrategia Nacional de Inteligencia Artificial 2026 ties this national artificial intelligence strategy to Bitcoin innovation as a growth engine. ANIA calls the plan living, built to shift as the sector matures.
Government of North Macedonia — North Macedonia became the 46th signatory to the Council of Europe Framework Convention on Artificial Intelligence, the first legally binding international treaty on this technology. Justice Minister Igor Filkov and Secretary General Alain Berset signed it in Skopje, marking North Macedonia AI regulation progress meant to keep artificial intelligence systems accountable to democratic norms and legal safeguards, part of a wider AI governance treaty push.
Government of Cabo Verde — Cabo Verde AI policy now runs through Resolution No. 72/2026, which the Government cleared via the Council of Ministers to lay out Agenda Digital 2035 as the country's blueprint for weaving AI, data-driven industry and automation into state functions. The resolution sets eight strategic axes meant to push the archipelago toward becoming a forward-looking, adaptable digital nation by 2035, spanning economic growth, e-government, social equity and green sustainability. The broader push for AI-driven digital transformation sits within this wider national plan rather than as a separate initiative.
Office of the Chief Justice — Clerical chores yes, rulings no: that's the line South Africa's Office of the Chief Justice draws in its new AI policy for the judiciary, the Draft Policy on Artificial Intelligence. Generative AI may assist with scheduling, case-flow tracking, summarising documents and translating, but stays away from weighing evidence, judging witness honesty, reading statutes, or fixing bail, sentences and risk scores. Anyone relying on it must disclose that, cross-check against trustworthy sources, finish risk training beforehand, and face review at least every two years.
Agencija za zaštitu osobnih podataka (AZOP) — AI data protection now has a formal reference point in Croatia: on 7 May 2026, the Personal Data Protection Agency (AZOP) issued guidance covering anyone who builds, trains, tests, integrates or runs AI systems. Titled "Data Protection and Artificial Intelligence", it clarifies GDPR applicability to AI models, assigns controller and processor responsibility, and details lawful processing grounds, privacy-by-design duties, transparency requirements, DPIA thresholds and security measures, alongside model anonymity, LLM risk factors and privacy-enhancing technologies. Organisations gain one benchmark for data protection and AI compliance across the development lifecycle.
Ministerio de la Presidencia de Panamá — Panama AI policy gains a permanent Artificial Intelligence Subcommission, led by SENACYT's National Secretary, via Executive Decree No. 36. A National Commission for Critical and Emerging Technologies must bring a National AI Strategy before the Cabinet Council for sign-off. Gazetted 8 May 2026 (No. 30520-A), this AI governance measure deems critical and emerging technologies strategic priorities.
Government of the Republic of Congo — A half-year deadline now governs eleven strategic outputs due from Republic of Congo's newly installed technical panel, tasked with shaping the country's national AI strategy. Frédéric Nzé heads the group, backed by the United Nations Development Programme. Aims include a Lingala and Kituba language dataset for local-language AI tools, costed at 150 to 200 million FCFA as part of wider artificial intelligence policy development.
Instituto Nacional de Tecnologias de Informação e Comunicação (INTIC) — INTIC opened public consultation on version one of the Proposal for the National AI Strategy, a Mozambique AI policy draft linking AI to digital transformation. Priority pillars span health and science, innovation and infrastructure, governance and public services, education and skills, entrepreneurship and economy, climate and agriculture, energy and environmental impact and culture and tourism. Comments closed 4 September 2026; the national AI strategy awaits adoption.
Government of French Polynesia (Conseil des ministres) — The Government of French Polynesia's Comité consultatif Smart Polynesia opened 18 June 2026 under President Moetai Brotherson, steering French Polynesia AI policy, digital sovereignty and network access. Members: the Haut-commissaire, ministers, DGEN, DSI, UPF, CCISM, OPEN, French Tech, MEDEF, CPME — spanning data hosting, skills and jobs, backing the wider plan for artificial intelligence governance.
Government of Mozambique (INTIC) — A draft Mozambique AI policy has been thrown open for public comment by INTIC, Mozambique's National Institute of Information and Communication Technologies. Universities, firms, government bodies and ordinary citizens can respond to the National Artificial Intelligence Strategy, pitched as a means to lift innovation, productivity and the quality of public services.
Sultan of Oman — Sultan Haitham bin Tarik's Royal Decree 50/2026 creates an AI special zone, the Artificial Intelligence Special Zone, in Muscat Governorate, granting resident firms exemptions, incentives and facilities under Oman's Law of Special Economic Zones and Free Zones. The Public Authority for Special Economic Zones and Free Zones oversees this AI free zone with the Ministry of Transport, Communications and Information Technology.
Ministry of Economic Development and Digitalization — The Ministry of Economic Development and Digitalization and the Ministry of Education and Research formed the Sub-Council for Artificial Intelligence and Applied Research (SCIA), an advisory body for Moldova AI policy. Joining academia, public authorities and the private sector, SCIA steers EU AI Act transposition into domestic law and names an authority to implement national AI governance.
National Judicial (Policy Making) Committee — Pakistan's National Judicial (Policy Making) Committee has issued the National Guidelines for the Use of Artificial Intelligence in Judicial Institutions, approved at its 57th meeting. They let AI in the courts help with legal research, document processing, case management and other functions, assisting rather than replacing judicial decision-making and human judgment.
Presidencia del Consejo de Ministros — Peru's Presidencia del Consejo de Ministros has signed off on a Peru AI policy running to 2030: the Estrategia Nacional de Inteligencia Artificial (ENIA) 2026-2030, approved via Resolución Ministerial N.° 152-2026-PCM. Four pillars anchor this national AI strategy — growing AI talent and capabilities, driving AI-based innovation and entrepreneurship, building an ethical and regulatory framework, and encouraging citizen participation and collaboration. Government, academia, private business and civil society are all meant to follow its direction on how artificial intelligence gets built, used and overseen nationwide.
UK Parliament — The Crime and Policing Act 2026, now law after Royal Assent, lets the Secretary of State rewrite the Online Safety Act 2023 to close AI regulation gaps around illegal AI-generated content and services enabling priority offences. Producing fake nude or sexualised images with AI tools becomes its own crime under this AI image generation law.
Ministry of Economic Development and Digitalization — Public feedback is being gathered in Moldova on a draft National Programme on Artificial Intelligence and Data Management, covering the years 2026 to 2030 and led by the Ministry of Economic Development and Digitalization. The Moldova AI policy seeks to turn the country from a mere buyer of foreign technology into a builder of homegrown AI products, bringing data oversight rules closer to EU standards along the way. This broader national AI strategy signals an effort to grow domestic tech capacity rather than depend on imports.
Cabinet of Iraq / Prime Minister's Office — Iraq AI policy now centres on a cabinet-backed roadmap covering 2024 to 2030, revealed at the Baghdad Tech Forum with USD 250 million earmarked over four years. The Iraqi Artificial Intelligence Group, under the Prime Minister's Office, drafted it around oil, public banking, healthcare, agriculture and Arabic-language government services. By 2028, 60% of government AI processing must run on servers inside Iraq.
Ministry of Communication, Digital Technology and Innovations — A decade-long Ghana AI policy took effect with Cabinet backing: the Republic of Ghana National Artificial Intelligence Strategy 2025-2035, unveiled 24 April 2026 by Minister Samuel Nartey George's Ministry of Communication, Digital Technology and Innovations. Funding comes via a National AI Fund rising from GHS 5 billion to GHS 15 billion, aimed at fair, ethical uptake spanning education, health, farming, banking and government services. Workforce skills, data rules and infrastructure spending anchor the national artificial intelligence strategy's foundations.
Ministry of Communication, Digital Technology and Innovations — Ghana is drafting an Emerging Technologies Bill, a step toward formal AI regulation for rolling out artificial intelligence alongside other advanced technologies. Minister Samuel Nartey George revealed the plan at the National AI Strategy launch, calling it the legal backing needed for safe, effective use while curbing risks. The bill sits within a broader legislative overhaul and will pair with the incoming Data Protection Bill for fuller AI oversight nationwide.
Ministério da Gestão e da Inovação em Serviços Públicos (MGI) — Brazil AI governance policy now covers MGI, the Ministry of Management and Innovation in Public Services, via Portaria 3.485, fixing principles, guidelines and responsibilities for developing, acquiring and using AI systems inside the ministry. It functions as internal AI regulation for public administration rather than external rulemaking.
Ministry of Finance, Budget and Digital Economy / UNDP — Burundi's Ministry of Finance, Budget and Digital Economy, with the United Nations Development Programme, cleared the Stratégie nationale de l'intelligence artificielle, or SNIA, on 21 and 22 April 2026. Six pillars carry the Burundi AI policy: public leadership, data plumbing, schooling, invention, high-impact uses, ethics, inclusion. The national AI roadmap promises 1000-plus state employees trained, fifteen trial schemes in agriculture and health, incubators at Bujumbura, Ngozi and Gitega, and by 2030 a driver of fair growth and youth work.
Ministry of Finance, Budget and Digital Economy — Burundi AI strategy has cleared final approval: the Ministry of Finance, Budget and Digital Economy, backed by the UNDP, signed off on the Stratégie nationale sur l'intelligence artificielle in Bujumbura. Running 2025-2030, six pillars underpin it — oversight via a new National AI Commission and ethics code, data systems, workforce skills (1,000-plus officials trained), entrepreneurship, farming and health trials, and sustainability. Government and priority sectors nationwide will follow this roadmap for AI adoption.
Ministerio de Tecnologías de la Información y las Comunicaciones (MinTIC) — MinTIC has rolled out updated security and privacy guidelines for artificial intelligence systems in Colombia, revising the Modelo de Seguridad y Privacidad de la Información (MSPI) under new AI governance in the public sector rules called Lineamientos de Seguridad y Privacidad de la Información para Sistemas de Inteligencia Artificial. These set technical, organisational and regulatory standards for building and maintaining AI tools inside government bodies.
Judicial Yuan — Taiwan's Judicial Yuan has issued judicial AI guidelines fixing 18 rules for AI use in courts and subordinate agencies. The AI Systems Reference Guidelines cast AI purely as an assistive device, bar confidential case files from unvetted platforms, and leave sentencing, fact-finding and evidence weighing to human judgement alone.
Republic of Azerbaijan — Azerbaijan has rewritten its Criminal Code, Criminal Procedure Code and its media and information statutes to reach artificial intelligence. Deepfakes now anchor Azerbaijan AI regulation: new article 148-2 makes it an offence to build or circulate unreal photo, video or audio drawn from someone's face or voice without permission, while article 242-1 does the same for pornographic or sexually explicit output. Machine-made matter shown publicly must bear a label that is easy to spot under the new AI law.
Ministry of Interior and Administration of Poland — Poland's Ministry of Interior and Administration has created a Team for the Application of Artificial Intelligence Systems (Zespół AI), spanning Police, State Fire Service, Border Guard, State Protection Service and Office for Foreigners. Its brief for AI in national security covers biometric identification, emergency-call categorisation, critical infrastructure prediction and large language model applications, plus AI for border control and migration management. A full strategy is due by 2035.
Government of Maldives — Maldives will submit a Maldives AI regulation bill to Parliament, President Dr Mohamed Muizzu confirmed, packaged alongside a Digital Transformation Bill, Digital Identity Bill and Cyber Security Bill covering artificial intelligence oversight. The four measures anchor the Maldives 2.0 digital transformation agenda, which targets full state-system digitisation by 2027. This push forms one strand of a wider digital-governance overhaul, not a standalone law.
Dutch government — Ten sitting watchdogs, not one new body, will handle Netherlands AI Act implementing legislation under the Uitvoeringswet verordening artificiële intelligentie, open for consultation. These include the Dutch Data Protection Authority (AP) and the Authority for Digital Infrastructure (RDI), which also jointly steer a cross-sector AI regulatory sandbox proposed in the same bill.
National People's Assembly deputies of Algeria — Algeria AI regulation has reached parliament: APN deputies filed the country's first AI legislative proposal, awaiting Bureau review. The bill would govern AI development and use to protect citizens and spur innovation, widen the ANPDP's oversight to AI systems, and extend privacy rules to AI-specific risks like profiling and automated decision-making, plus a legal basis for public-sector AI procurement.
Ministry of Electronics and Information Technology (MeitY) — A fresh advisory layer now underpins AI governance India relies on: MeitY (Ministry of Electronics and Information Technology) has launched the Technology and Policy Expert Committee (TPEC). Scholars, corporate experts and digital-policy specialists staff it, converting technical matters into direction for the AI Governance and Economic Group, reinforcing national AI policy and regulation.
Ministry of Electronics and Information Technology (MeitY) — India's AI regulation now runs through a new inter-ministerial panel, the AI Governance and Economic Group (AIGEG), launched by the Ministry of Electronics and Information Technology under Ashwini Vaishnaw. AIGEG steers AI governance policy nationally, putting proposals from the India AI Governance Guidelines and the Economic Survey into practice, backed by a Technology and Policy Expert Committee.
UK Parliament — New rules under the Data Protection Act 2018 require the Information Commissioner to produce a statutory Code of Practice on AI regulation compliance, addressing how organisations handle personal data while building and deploying Artificial Intelligence and automated decision-making tools, once regulations took effect on 12 May 2026. Special attention must go to safeguarding children's personal data within these automated systems. The outcome: formal UK data protection guidance on AI use.
Ministry of Economic Development and Digitalization — Moldova's Ministry of Economic Development and Digitalization and the e-Government Agency have drafted a bill legalising AI in public services from 1 January 2027, provided humans oversee outcomes. Fully automated rulings stay barred for sensitive cases like suspending welfare payments or settling child custody, though citizens gain appeal rights. This tightens safeguards on AI in government administration.
Regierung des Fürstentums Liechtenstein (Government of the Principality of Liechtenstein) — Public-sector AI use in Liechtenstein now runs on four pillars: accountability, putting people first, openness, and innovation-driven efficiency. The recently approved KI-Strategie der Liechtensteinischen Landesverwaltung (AI Strategy of the Liechtenstein State Administration) marks an AI strategy for public administration, structured around four action fields: people, organisation, governance, and infrastructure, aimed at steering AI adoption in government agencies. Officials set no fixed end date, but a fundamental revision must happen no later than 2030.
National Internet Information Office, National Development and Reform Commission, Ministry of Industry and Information Technology, Ministry of Public Security, State Administration for Market Regulation — China's Cyberspace Administration (CAC) leads four ministries behind Order No. 21, curbing AI companion apps that imitate human personality and speech for extended emotional bonding. Minors cannot get virtual intimate relationship features, platforms above 1 million registered users face mandatory safety review, and a regulatory sandbox now applies under the Interim Measures for the Management of AI Personified Interactive Services. Force began 15 July 2026.
Ministry of Information Technology, Communication and Innovation / AI Unit — Mauritius's Ministry of Information Technology, Communication and Innovation unveiled the FAIR Guidelines on 10 April 2026, pairing them with the National AI Strategy to shape Mauritius AI governance nationwide. Four pillars—Fairness, Accountability, Inclusiveness and Integrity, and Responsibility—anchor this voluntary code, steering responsible artificial intelligence adoption, design and deployment across industries. Regulators intend the code to harden into binding rules once the sector matures.
Corte Suprema de Justicia del Paraguay — Paraguay's Supreme Court has approved Resolution No. 12677, AI guidelines for the judiciary covering judges and staff. AI must support judicial staff without replacing human input in decisions, courts must report judicial AI use where rights may be affected, and UNESCO's Montevideo regional office gave technical cooperation.
National Assembly of Congo — The National Assembly of Congo passed a bill steering scientific research and technological innovation, a text carried by the Ministry of Scientific Research and Technological Innovation to modernise the sector and absorb what artificial intelligence now demands. This Congo AI regulation writes an ethical framework around such technologies and stands up a national committee accountable for responsible use, answering to norms in force and to social values. Caria is also foreseen to organise work in the field. Jean-Luc Marcel Mouthou, whose portfolio covers preschool, primary and secondary education plus literacy, presented these AI governance rules to deputies, calling them an important step.
Government of the Slovak Republic — Slovakia has created the Rada vlády SR pre digitálnu transformáciu a umelú inteligenciu (Government Council for Digital Transformation and AI), on a MIRRI proposal. It replaces the former Government Council for Digitalisation of Public Administration and the Single Digital Market, steering Slovak AI policy and digital transformation through coordinating work spanning AI governance coordination government-wide.
Government of Jordan — Jordan's Regulation No. 13 of 2026, the Digital Media Regulation Bylaw, now imposes AI disclosure in digital media on every licensed operator. Article 18 bars AI-generated fake, misleading, inciting or hateful output and requires consent before personal data trains AI systems. Binding from 23 May 2026 under Media Commission oversight, following Council of Ministers approval on 5 April and Official Gazette publication on 23 April, it is Jordan's first such AI transparency rule.
Saudi Data and Artificial Intelligence Authority (SDAIA) — Saudi Arabia's SDAIA has floated a draft Responsible AI Policy for public comment, reaching state and private bodies via Istitlaa, due 3 May 2026. This Saudi AI regulation grades systems into four bands, critical to low, forcing registration, ethics labels and audits on the riskiest, plus new inspection and data rules. Feedback closed, it becomes a binding Saudi AI governance framework.
Government of Estonia — Ministers approved and passed to parliament a bill setting a technology-neutral legal frame around automatic administrative procedure — Estonia AI regulation reaching routine, repeated and data-driven rulings. Central and local bodies would gain authority to decide through algorithms and machine learning, weighed against citizens' constitutional guarantees. Where an automated route proves opaque or ill-suited, the applicant may turn to a person, a safeguard inside the AI rules for administrative decisions.
Ministry of ICT and National Guidance — Uganda AI strategy work is nearing completion under the Ministry of ICT and National Guidance, which is finalising a National AI and Emerging Technologies Strategy, an AI and emerging technologies policy meant to steer artificial intelligence, blockchain, IoT and robotics uptake across sectors. Nationwide consultations closed in March 2026, with sign-off expected by June 2026 and agriculture, health, education and public administration named as early focus areas. Ethics and oversight provisions will follow Uganda's Data Protection and Privacy Act.
Macao SAR Government — Wong Sio Chak says Macao plans to avoid heavy-handed Macao AI regulation, fearing rules could stifle innovation. Legal frameworks for artificial intelligence are still being worked out globally by trial and error, he argued, so strict rules now risk doing more harm than good. Macao has no AI-specific legislation yet, unlike mainland China, which is racing to pass its own AI governance framework.
Rada Ministrów (Council of Ministers of Poland) — AI regulation Poland now covers content labelling, a public registry for sanctioned or withdrawn systems, and legal grounds for AI regulatory sandboxes, following approval by Poland's Council of Ministers. Officials call it the EU's first such wide-ranging AI policy framework, with the draft AI Systems Act next heading to parliament.
Ministry of Economy, Trade and Industry / Ministry of Internal Affairs and Communications — AI governance Japan just got its latest update: METI and MIC issued Ver1.2 of the AI Guidelines for Business on 31 March 2026. Building on Ver1.0 (April 2024) and Ver1.1 (April 2025), the revision spells out shared principles so business actors can spot dangers and adopt voluntary safeguards throughout an AI system's lifecycle, while pushing safer, more secure use nationwide. New in Ver1.2: formal definitions of AI agents and physical AI, plus a mandate that human judgment be built into design from the outset. The move reflects Japan's wider push on AI regulation, favouring lifecycle oversight over rigid rulebooks.
Receita Federal do Brasil — Brazil's Federal Revenue Service presented its AI policy for tax administration and customs, with ethical, legal and technical rules, at a UN meeting on 27 March. Final decisions and legal acts stay with civil servants, misuse is a disciplinary matter, mass surveillance, social scoring of citizens and subliminal manipulation are banned, and a curator monitors generative-AI errors and bias.
Israel Judicial Authority — Israel's Supreme Court president, Yitzhak Amit, and courts director Tzachi Uziel have issued an AI ethics code for courts to every judge, registrar and staff member. Prepared by a recently appointed team, the judicial AI rules set principles for responsible use across the courts' work and will be reviewed and updated periodically.
Corte Suprema de Justicia / Organismo Judicial — A new 32-article rule confines AI use by judges in Guatemala to assistive, administrative roles only. Acuerdo de Presidencia 93-2026 bars the Corte Suprema de Justicia from letting algorithms generate rulings alone, leaving evidence weighing and verdicts strictly to magistrates under this judicial AI regulation.
Supreme Court of Israel — Israel's Supreme Court made Ramat Gan municipality pay a resident NIS 30,000 in costs after AI-fabricated citations tainted its court filings and its school-transport decision for a special-education pupil. Justice Noam Sohlberg noted the erroneous, partly invented, likely AI-generated references recurred three or four times, even after the court raised them, and said uncontrolled AI use by a public authority toward citizens is especially serious. The municipality says it will learn from this and pursue the lawyer responsible.
Department of Information and Communications Technology — The Department of Information and Communications Technology opened consultation on its Draft National Sovereign Digital Transformation and Artificial Intelligence Strategy, placing Papua New Guinea AI policy, digital government and digital public infrastructure under one sovereign architecture. Four laws are proposed: a National Artificial Intelligence Act, Data Governance and Protection Act, Digital Identity and Trust Framework Act and Critical Digital Infrastructure Investment Act. New bodies would follow: a Digital Transformation and AI Board, a restructured Department of Digital Transformation and Artificial Intelligence and a National Digital Economy Authority. None exists yet, because this AI governance plan is still a draft.
Government of Luxembourg — Luxembourg's Council of Government now requires human review of AI in central administration before content is published, holding the approving civil servant accountable rather than the system itself. Disclosure duties also apply, matching Article 50 of the EU AI Act, part of wider AI governance in public administration.
Parliament of Chad (National Assembly and Senate) — Chad places artificial intelligence under ANSICE's formal AI oversight through Law No. 003/AN-SENAT/2026, promulgated by Decree No. 148/PR/2026, ratifying Ordinance No. 001/PR/2026. The reform extends the national agency beyond information security and electronic certification into an expanded AI regulatory field, folding in digital platforms, social media and AI systems, while sharpening its powers over cybercrime, data protection and cybersecurity.
Ministry of Digital Transformation — Slovenia's Ministry of Digital Transformation has adopted the Guidelines for the Development, Introduction and Use of Artificial Intelligence in Public Administration, a framework meant to steer government bodies toward deploying AI in public administration safely, ethically and effectively when serving citizens and businesses. Rather than binding rules, officials describe it as offering practical grounding for weaving automated systems into routine state functions — part of a broader push toward artificial intelligence in government services. Two attached annexes fill in specifics: guidance for staff on handling publicly available generative tools, plus an exit plan for cloud providers spelling out mandatory security benchmarks.
Ministry of Digital Technologies of the Republic of Uzbekistan — Uzbekistan's Minister of Digital Technologies has issued an AI governance framework under Order No. 284-mh, registered by the Ministry of Justice as No. 3787: the Ethical Rules for Creating, Implementing and Using Artificial Intelligence-Based Solutions. It binds developers, implementers and users, defining AI and AI systems while requiring safeguards for human rights, freedoms and legitimate interests. Uzbekistan's AI ethics standards take legal effect from 17 June 2026.
National Council for Artificial Intelligence (NCAI) — Egypt's National Council for Artificial Intelligence, Quantum Computing, and Emerging Technologies published its national generative AI policy on 14 March 2026, a principles-based framework spanning five categories of users, from developers to everyday individuals. These generative AI guidelines tackle hallucination, misinformation, synthetic media, bias, privacy and intellectual property risks, with dedicated rules for agentic AI, deepfakes, classroom use and disclosure duties. The document is built as a living guide meant for periodic revision, not a fixed rulebook.
ChileCompra (Dirección de Compras y Contratación Pública) — Under Directiva de Contratación Pública N° 44/2026, Chile's ChileCompra agency guides public buyers toward safer AI procurement. Acquiring data science and artificial intelligence tools through Mercado Público now requires assessing dangers such as skewed algorithms, poor transparency and cybersecurity gaps, then building multidisciplinary teams, running ethical reviews and embedding audit clauses in contracts — advancing government artificial intelligence contracting.
House of Representatives — Deepfake regulation arrived in Cyprus through Law 29(I)/2026, revising the Copyright and Related Rights Law against unauthorised synthetic media. Producing or sharing deepfakes without consent now carries criminal liability plus civil remedies, with fabricated likeness safeguards lasting 50 years past a person's death.
Government of Romania — Four separate bodies, not a single regulator, will police EU AI Act Romania compliance under a Memorandum passed by the country's cabinet. ANCOM leads as the designated contact point and general market watchdog, ASF and BNR govern high-risk AI in financial services, ANSPDCP handles biometric technology and law-enforcement applications, and ADR vets conformity assessment bodies. Regulation 2024/1689 thus takes effect through this layered domestic structure rather than a unified agency.
National Council for Artificial Intelligence (NCAI) — Four risk tiers, prohibited through minimal, now classify AI systems in Egypt under a fresh National AI Governance Framework cleared by the National Council for Artificial Intelligence. Five ethical principles underpin the rules, paired with a dual-check process for high-risk tools and separate rules for general-purpose AI models, building toward binding legislation ahead. Guidance on trusted, responsible and generative AI regulation followed simultaneously.
Ministry of Digital and CyberSecurity Malaysia — One national hub now anchors AI standards Malaysia: MY-AI Standards, engineered by CyberSecurity Malaysia. Gobind Singh Deo unveiled the site, with support from the National AI Office, the Department of Standards Malaysia and SIRIM Bhd, calling it trust infrastructure for reaching accepted global benchmarks. Built around an AI governance framework, the hub bundles guidance, cooperation and backing while countering deepfake dangers.
Ministry of Science and Technology (Bộ Khoa học và Công nghệ) — Vietnam's Ministry of Science and Technology has promulgated the National AI Ethics Framework through Circular 05/2026/TT-BKHCN, effective immediately. This Vietnam AI regulation binds artificial intelligence research, development and deployment to core principles: human-centred design, transparency, fairness, non-discrimination and accountability. Modelled partly on UNESCO's guidance, the framework gives organisations and individuals a reference point for lawful AI governance practices.
Cayman Islands Civil Service — The Cayman Islands Civil Service has completed a first Artificial Intelligence Policy (Version 1.0), Premier André Ebanks told Parliament, the Office of the Deputy Governor yet to sign off the Cayman Islands AI policy Government-wide. Risk management, governance, transparency, accountability and oversight duties span the service; procurement, piloting or deployment of any AI system awaits joint review of legal compliance, cybersecurity, data protection, architectural alignment and risk assessment. These AI governance rules carry enforcement and bar unapproved deployment, sensitive-data misuse, unlawful or harmful use and high-risk applications; Ebanks targets Q2 2027 for draft AI legislation.
Cayman Islands Government — Cayman Islands AI legislation should reach draft stage by Q2 2027, Premier André Ebanks told Parliament, previewing an AI legislative framework for public-sector AI use. A taskforce set up October 2025 within the Premier's office and the Ministry of Social Development and Innovation owes policy recommendations by end-May 2026 to shape the drafting. Ebanks also flagged a near-finished Civil Service AI Policy (v1.0), locking in mandatory governance, risk and oversight duties for government AI tools.
Government of the Republic of Slovenia — Five pillars anchor the national AI strategy of Slovenia, set by the Government of the Republic of Slovenia as NsUI 2030 on 5 March 2026 for the period to 2030: sovereign, trustworthy AI; broader uptake across business and government; domestic AI skills and ecosystem growth; defence and security uses; and stronger global ties. This Slovenia AI policy supersedes the earlier National Programme for AI by 2025.
Ministry of Interior (Iraq) — Iraq's Ministry of Interior has issued 'Controls and Ethics of AI Use in the Ministry of Interior – 2026', AI ethics rules drafted by the ministry's cybersecurity directorate on ministerial orders. These AI use controls rest on national AI ethics from a specialist government team and international data-protection and human-rights standards, to strengthen digital security.
Informacijski pooblaščenec (Information Commissioner of the Republic of Slovenia) — Slovenia's Information Commissioner issued guidance on responsible AI use in government, covering when personal data may enter AI tools and when an individual counts as non-identifiable. It requires AI answers to be verified and confirms decisions stay with a human official, tying safe AI adoption in the public sector to data protection rather than the tools themselves.
Ministry of Communication, Digital Technology and Innovations — Samuel Nartey George, Ghana's Communications Minister, unveiled a draft law for AI regulation Ghana needs, covering automated decision-making, deepfakes, and data transfers across borders. Speaking at Accra's Data Protection Conference, he framed it as overhauling enforcement as AI spreads across health, finance, telecoms and government. A parallel Emerging Technologies Bill would extend AI systems oversight to advanced analytics and digital assets.
Council of Ministers of Ethiopia — Ethiopia's Council of Ministers cleared rules creating the Medemer Artificial Intelligence University, a self-governing state campus in Addis Ababa for AI higher education and advanced computational sciences. The measure fixes its leadership, independence and teaching remit. Under Digital Ethiopia 2030, its AI tertiary education prepares graduates in robotics, machine learning, advanced computing and data science.
Government of Gabon — Gabon's Ordinance n°0011/PR/2026, dated 26 February 2026, introduces AI content labelling: synthetic text, images, video and deepfakes must bear an indelible mark showing they are artificial. Articles 32 to 34 and 53 also oblige platforms to run automatic detection systems, and outlaw explicit sexual deepfakes, wrongly crediting statements to public figures, and posing as state institutions. The rules rank among Africa's earliest binding measures on synthetic media marking.
Government of Moldova — Moldova has paused Moldovan AI law drafting on EU advice, said Michelle Iliev, State Secretary at the Ministry of Economic Development and Digitalization. The European Commission is simplifying its own AI Act via the Digital Omnibus package, making a parallel effort unnecessary. AI regulation resumes once the EU framework is finalised.
National AI Strategy Committee — South Korea's National AI Strategy Committee (국가인공지능전략위원회) passed a three-year national AI regulation blueprint: 99 implementation tasks, 326 policy recommendations, under Article 6 of the AI Basic Act. The 'Korea AI Action Plan (2026~2028)' (대한민국 인공지능행동계획) rests on three pillars — AI innovation ecosystem, nationwide AI-based transformation, global AI basic society — across 12 domains, including infrastructure, industry and public AX, talent, and global AI governance policy.
Gibraltar Regulatory Authority — The Gibraltar Regulatory Authority, acting as Information Commissioner, is among 61 signatories to a Joint Statement on AI-Generated Imagery, convened by the Global Privacy Assembly. Signing sets Gibraltar AI regulation alongside a common line on artificial intelligence that fabricates convincing photographs and video of identifiable people who never consented, children most at risk. Builders and users must install AI imagery safeguards, be candid about capability, pull harmful material on request, and shield children.
Office of the Data Protection Authority — Guernsey's Office of the Data Protection Authority is among 61 regulators behind the Joint Statement on AI-Generated Imagery and the Protection of Privacy, which the ODPA convened as secretariat and co-chair of the Global Privacy Assembly's International Enforcement Cooperation Working Group. Guernsey AI regulation now bites on systems fabricating lifelike pictures and video of real people who never agreed, minors and vulnerable groups most at risk, and co-signatories want strong safeguards, candour about system capability, reachable takedown routes for damaging material, and added protection for children. Parallel AI image rules form a Crown Dependency Data Protection Advisory on generative AI image creation, co-issued with the Isle of Man Information Commissioner and Jersey's Office of the Information Commissioner, restating that indecent images of a child stay criminal, AI-made or not, compliance resting with whoever builds and uses them. Islanders are told to tighten social-media privacy settings, treat AI apps wanting photographs or personal data warily, grasp their entitlements, and discuss internet safety with children.
Gibraltar Regulatory Authority — The Gibraltar Regulatory Authority co-signed a Joint Statement on AI-Generated Imagery with 60 other data protection authorities on 23 February 2026, addressing AI image generation privacy risks from realistic images and videos of identifiable individuals made without consent. It urges safeguards, transparency, removal mechanisms, and protections for children against AI-generated imagery of identifiable individuals.
Nigeria Data Protection Commission — Nearly 60 privacy regulators worldwide joined Nigeria's Data Protection Commission in signing the Joint Statement on AI-Generated Imagery and the Protection of Privacy. Compliance Audit Returns under the Nigeria Data Protection Act, plus the General Application and Implementation Directive, will now judge responsible AI use.
Corte Suprema de Justicia (Supreme Court of Justice) — Human judgment cannot be delegated away, per new guidance from Costa Rica's Supreme Court of Justice, issued through its Superior Council on 20 February 2026 for generative AI in judicial work. The signer of any ruling bears full responsibility, states the circular 'Lineamientos básicos para el uso de la Inteligencia Artificial Generativa autorizada en el Poder Judicial', binding court staff, the Public Prosecutor's Office, the Public Defence and the Judicial Investigation Agency to treat such AI use in courts as support, not substitution. Personal data must be anonymised manually, and staff must disclose which AI tool was used and how.
Regeringskansliet (Finansdepartementet) — Sweden's new AI policy, Sveriges AI-strategi, supersedes the 2018 National Approach. It targets a top-ten global ranking for artificial intelligence and world-leading AI adoption for public administration.
Ministry of Popular Power for Science and Technology (Mincyt) — Nine principles now anchor AI regulation Venezuela, among them humanist AI, equity, safety and environmental responsibility. This new AI ethics framework, unveiled by Venezuela's Ministry of Science and Technology, is called the Code of Ethics for the Development and Responsible Application of Artificial Intelligence, meant to steer tech use on ethical grounds. A related law has cleared its first National Assembly reading.
Communications and Information Technology Regulatory Authority (CITRA) — CITRA opened feedback on a proposal called "Guidelines for Adopting AI in the State of Kuwait," reaching across every industry. Kuwait AI regulation weighs moral safeguards, privacy protections and alignment with the country's growth agenda, accepting responses through 20 April 2026. It marks CITRA's debut national AI governance framework.
Senate of Kenya — Kenya's Senate has tabled the Artificial Intelligence Bill, 2026, entering Kenya Law's Bills collection. The proposed Kenya AI regulation, a 1.1 MB PDF, remains under Senate consideration with no passage timeline, marking an early step toward AI governance rules.
Communication and Information Technology Regulatory Authority (CITRA) — CITRA released the Guide to Adopting Artificial Intelligence in the State of Kuwait, version 1.0, setting out how businesses and government bodies should responsibly pursue AI adoption in Kuwait. The document addresses governance structures, data handling, infrastructure needs and workforce skills, tying these to Kuwait Vision 2035, as part of the country's broader approach to AI adoption. A public feedback period, closing 20 April 2026, has now ended.
Supreme Court of the Philippines — The Supreme Court of the Philippines set out a Governance Framework on the Use of Human-Centered Augmented Intelligence in the Judiciary in a resolution dated 18 February 2026. It lets AI in the judiciary help with court documents, from copy-editing, optical character recognition and summarising to legal research, translation and voice-to-text transcription, but bans any tool that could harm stakeholders, breach rights or weaken the rule of law. AI output can never be the sole basis for adjudication, and human decision-makers keep full responsibility for legal reasoning and judgments.
Ministry of Information and Communications Technology — The Ministry of Information and Communications Technology's Department of E-Government opened validation of the Draft National Artificial Intelligence Strategy at a two-day Lilongwe workshop. This Malawi AI policy sets a coordinated framework for responsible AI adoption across ministries, plus stronger digital public infrastructure and innovation. By August 2026, the national AI plan remained unadopted, pending finalisation.
Personal Data Protection Committee (PDPC) — Thailand's PDPC (Personal Data Protection Committee) has released draft AI data protection guidelines, titled Guidelines on Personal Data Protection in AI Development and Use, translating Personal Data Protection Act duties for controllers and processors into AI-specific rules. The 17 February text names stakeholder roles, bars certain model training uses in data processing agreements, requires impact assessments for high-risk AI, and sets security measures across the AI lifecycle. Comments close 25 February 2026.
Cabinet of Mauritius / Ministry of Information Technology, Communication and Innovation — Mauritius now follows a national AI strategy Mauritius will use through 2029: "AI for Mauritius", Cabinet's guiding artificial intelligence framework economy-wide. Priority sectors span health, finance, farming, industry and government services, overseen by a new Mauritius Artificial Intelligence Council. It is the country's first unified AI policy framework Mauritius has set nationally.
Superintendencia de Protección de Datos Personales — The Superintendence of Personal Data Protection (SPDP) issued Resolución No. SPDP-SPD-2026-0009-R: 'Norma General para la Garantía del Derecho de Protección de Datos Personales en el Uso de Sistemas de Inteligencia Artificial'. This Ecuador AI regulation binds data controllers and processors handling local personal data with AI, setting principles, obligations and control mechanisms. The data-protection rules for AI mandate impact assessments, risk management, security measures and audits, and let the SPDP audit systems and order corrective measures.
Superintendencia de Protección de Datos Personales — Ecuador now has binding AI data protection rules reaching providers abroad whose systems touch residents' personal data. The Superintendencia de Protección de Datos Personales (SPDP) issued Resolution SPDP-SPD-2026-0009-R under the Norma General para la Garantía del Derecho de Protección de Datos Personales en el Uso de Sistemas de Inteligencia Artificial, an AI regulation splitting duties among developers, deployers, distributors and implementers across a system's full life: risk checks, disclosure requirements, and human review of automated outcomes.
Dirección Nacional de Vigilancia Sanitaria (Dinavisa) — A new Strategic Work Unit now guides AI regulation in health surveillance at Paraguay's Dinavisa, formed under Resolution 047/2026. Officials there will produce guidelines, weigh potential hazards against opportunities, and float pilot schemes, keeping artificial intelligence in regulatory processes strictly advisory rather than a substitute for human decision-making.
Ministry of Electronics and Information Technology (MeitY) — The Ministry of Electronics and Information Technology's Information Technology (Intermediary Guidelines and Digital Media Ethics Code) Amendment Rules, 2026, in force from 20 February 2026, are the first binding India AI regulation. Defining synthetically generated information, the rules oblige intermediaries to technically block unlawful AI-generated content, visibly flag other synthetic content and embed permanent, uniquely identified metadata. Significant social media intermediaries must also verify user declarations under the AI-specific regime.
Ministry of Information Technology & Telecommunication — Nine principles frame Islamabad's push for Pakistan AI regulation, unveiled by the Ministry of Information Technology & Telecommunication at Indus AI Week. Ownership stays local under the Islamabad AI Declaration: private industry drives the work, backed by real capability, public confidence, accountable institutions, sound safeguards and careful delivery, with tangible results for citizens as the test. Any state deployment touching public rights or core government functions has to stay explainable, checkable and matched to the risk involved, forming the backbone of AI policy in Pakistan.
Government of Kyrgyzstan — Kyrgyzstan AI regulation under Chapter 23 of the Digital Code (articles 191-197) now covers every AI system deployed nationally. Operators must publish a self-run risk assessment, disclose AI-generated interactions and deepfakes to users, and, for high-risk systems, issue a public conformity declaration. Cabinet Resolution No. 770 (December 2025) approved the methodology behind these AI governance requirements.
National Development and Reform Commission, Ministry of Industry and Information Technology, Ministry of Housing and Urban-Rural Development, Ministry of Transport, Ministry of Water Resources, Ministry of Agriculture and Rural Affairs, Ministry of Commerce, State-owned Assets Supervision and Administration Commission — Nationwide AI in public procurement bidding is now compulsory under Implementation Opinions on Accelerating AI Application in the Bidding Sector (Fa Gai Fa Gui [2026] No. 195), issued jointly by China's National Development and Reform Commission (NDRC) and eight other agencies. Mandated functions span the whole tender process: screening bid documents for compliance, machine-assisted evaluation, spotting bid-rigging, joint supervision and handling complaints. Officials set 2026 and 2027 as milestone years for rolling these AI-driven procurement scenarios out everywhere.
Receita Federal do Brasil (RFB) — An AI governance policy tax authority mandate now binds Brazil's Federal Revenue Service across a system's full life, from acquisition to retirement. Human review must approve each automated outcome; social scoring and mass surveillance are forbidden; a new "curator of generative AI" watches for hallucinations, bias and errors. This artificial intelligence policy for public administration took formal shape as the Artificial Intelligence Policy, published in the DOU.
Autoriteit Persoonsgegevens (AP) — The Autoriteit Persoonsgegevens (AP), the Dutch Data Protection Authority, finalised its stance on Netherlands AI regulation for generative AI systems, setting conditions for safe, responsible use that respects fundamental rights. A 'values at work' scenario weighs innovation against safeguards for democracy and civil liberties. The guidance sketches future paths for generative artificial intelligence oversight rather than fixed rules.
Secretaría de Ciencia, Humanidades, Tecnología e Innovación and Agencia de Transformación Digital y Telecomunicaciones — SECIHTI (Secretaría de Ciencia, Humanidades, Tecnología e Innovación) and the ATDT (Agencia de Transformación Digital y Telecomunicaciones) have issued a voluntary code on AI ethics in Mexico, grounded in the ten Principles of Chapultepec. Its formal name, Declaration of Ethics and Good Practices for the Use and Development of Artificial Intelligence, describes non-binding standards for designing, deploying and assessing AI systems across public institutions, private firms and civil society actors. The stated focus spans rights protection, inclusion, accountability and benefit to society, an AI governance approach that carries no legal force.
General Council of the Judiciary (CGPJ) — Spain's General Council of the Judiciary (CGPJ) approved Instruction 2/2026 on 28 January 2026, setting AI rules for judges and magistrates, generative AI included. Judges may use only AI systems from the competent justice administrations or the CGPJ, and AI may never replace them in deciding cases. The Official State Gazette (BOE) then published it.
Executive Yuan — The Executive Yuan approved the Ten AI Initiatives Promotion Plan (2025-2028), a Taiwan AI policy targeting NT$7 trillion in output value by 2028. The National Development Council coordinates this national AI strategy, which spans digital infrastructure, key technologies and smart applications.
ICT Division, Government of Bangladesh — A tiered, risk-based Bangladesh AI policy draft covering 2026-2030 is now open for public comment via the ICT Division. Mass surveillance and social scoring are banned, high-risk systems must pass algorithmic impact assessments, and a permanent AI Act is due by 2028. Interim officials describe it as their second attempt at national AI regulation, joining sovereignty goals with growth incentives through a new AI Innovation Fund.
Consejo General del Poder Judicial (CGPJ) — Spanish courts face binding curbs on AI use by judges under Instruction 2/2026 from the General Council of the Judiciary (CGPJ), published in the BOE on 30 January 2026. Algorithmic tools must stay under human command, autonomous rulings are barred, and only CGPJ-vetted software may be deployed. Profiling, behavioural forecasting and specially protected personal data via judicial artificial intelligence tools stay off-limits.
Executive Yuan — Taiwan's Executive Yuan has greenlit a four-year AI national strategy Taiwan calls the AI New Ten Major Constructions Promotion Plan, backed by roughly NT$1900億元. The National Development Council leads delivery with the Ministry of Economic Affairs, MODA and the National Science and Technology Council, spanning sovereign AI, computing capacity and talent pipelines. This national AI development plan releases funding year by year rather than all at once.
Government of Armenia — Armenia AI regulation gained a treaty anchor on 27 January 2026, when Yerevan endorsed the Framework Convention on Artificial Intelligence, Human Rights, Democracy and the Rule of Law, the first global legal text shielding freedoms from the technology. Foreign Minister Ararat Mirzoyan signed in Strasbourg, taking signatories to seventeen. Ratifying binds domestic AI rules to protect rights and democratic norms without stifling innovation, via law or executive action.
Government of French Polynesia — French Polynesia now has a charter for AI use by government employees, covering research, note-taking, drafting and translation. Vice-President Minarii Galenon signed the rules, barring staff from entering sensitive data into unapproved tools. Binding decisions stay with people, not AI in the public sector, unless a human supervises.
Government Office (Riigikantselei) — Estonia formed the Eesti.ai Advisory Board to steer AI in government under Bolt founder Markus Villig, who chairs the panel. Answering straight to Prime Minister Kristen Michal, it supervises major artificial intelligence work spanning the whole economy, alongside other prominent Estonian tech entrepreneurs and experts.
Government of the Republic of Korea — South Korea's AI Basic Act, formally the Act on the Development of Artificial Intelligence and Establishment of Trust, came into force on 22 January 2026. Generative AI operators must label AI-made content, and high-impact AI operators must summarise training data and meaningfully explain results, including the main criteria and principles behind them. Foreign AI firms with no Korean office must appoint a local agent past set revenue or daily-user thresholds, and administrative fines wait out a one-year grace period.
National Assembly — South Korea created the Presidential Council on National Artificial Intelligence Strategy under Act No. 20676, formally titled the Framework Act on the Development of Artificial Intelligence and the Creation of a Foundation for Trust, also called the AI Basic Act. This South Korea AI regulation forces disclosure rules onto generative and high-impact systems, orders safety checks for high-performance models, and funds industry and research growth. Ten sensitive fields fall under its reach — healthcare, employment, credit and public safety among them — anchoring a wider AI governance framework signed into law 21 January 2025.
Oliy Majlis of the Republic of Uzbekistan — Law No. ZRU-1115 penalises unlawful AI processing of personal data with fines of 50–100 times the base calculation value, plus confiscation, by amending the Administrative Liability Code and the Law on Informatisation. This Uzbekistan AI regulation treats AI as technology mimicking human cognitive functions, such as self-learning and solution search. Its AI law provisions forbid AI-created information resources and systems from harming people or their rights, and ban legally significant decisions on human rights and freedoms that rely only on AI conclusions.
First Lady of Mauritania and Ministry of Education — Mauritania's First Lady, Dr Mariem Fadel Dah, unveiled 'Artificial Intelligence in the Service of the Republican School' in Nouakchott, a national push for AI in education grounded in ethics and sovereignty. Teachers gain sharper tech skills, lessons improve, and the nation stakes a claim as a regional model for safe classroom adoption. Public, private and academic groups share one structure to push it forward together.
Assemblée nationale de Côte d'Ivoire (National Assembly) — 24 bills reached Côte d'Ivoire's National Assembly at the start of its ordinary session, and artificial intelligence legislation stands out among them. Officials frame the draft Côte d'Ivoire AI regulation as a way to align law with emerging digital challenges under an AI legal framework, while advancing the nation's ambition to become a regional tech hub. Parliament has yet to approve the text, leaving it pending further debate.
Isle of Man Government, Department for Enterprise (Digital Isle of Man) — £1 million backs a freshly created National AI Office (NAIO) steering Isle of Man AI regulation, overseeing how artificial intelligence takes hold across the Island. NAIO merges scattered government and industry roles under one roof, giving guidance aimed at growth, better public services and stronger AI literacy. Built atop Activate AI, the 2024 programme, it hands the Island's artificial intelligence strategy a lasting institutional home.
House of Representatives and King of Morocco — Morocco's Organic Law No. 53.25, reworking Organic Law No. 27.11 on the House of Representatives, now lists "artificial intelligence tools" as a possible instrument for election crimes. Anyone who deploys such tools to circulate fabricated content or false news aimed at damaging candidates or electoral integrity risks two to five years behind bars under this AI election law. Deputies passed the measure on 1 December 2025, the Constitutional Court validated it on 24 December 2025, and Dahir No. 1.25.70 sealed the electoral integrity legislation into force.
Egyptian Cabinet — Egypt's Cabinet has widened the National Council for Artificial Intelligence's remit to cover quantum computing, biotechnology and other emerging fields, renaming it the National Council for Artificial Intelligence, Quantum Computing, and Emerging Technologies. This Egypt AI policy move backs startups and lifts global standing via new rules under the broader Egypt AI regulation drive. Dating to 2019, the council is chaired by Egypt's IT minister.
Government of Taiwan — President Lai Ching-te promulgated Taiwan's AI Basic Act, in force since 14 January 2026 after the Legislative Yuan passed it on 23 December 2025. The principle-based Taiwan AI regulation names the National Science and Technology Council as central competent authority and has agencies help their sectors write AI safety guidelines and codes of practice. It also tells government to use AI to protect workers' rights and tackle skills mismatches.
Legislative Yuan and President of Taiwan — The Legislative Yuan approved the 20-article Artificial Intelligence Basic Act on 23 December 2025, anchoring Taiwan AI regulation. The statute makes the National Science and Technology Council central competent authority, fixes seven governance principles and adds cross-agency coordination. Sector regulators get two years to write the AI law's implementing rules.
Government of Indonesia — A pair of drafts folded into Presidential Decree 38/2025 (Perpres), inked by President Prabowo on 22 December, now anchors Indonesia AI policy. Guiding ethical standards alongside a tiered risk framework mirroring the EU AI Act make up one text; a National AI Roadmap mapping out 2026-2030 across 10 focus sectors forms the second. Every minister had endorsed the ethics draft by September 2026, awaiting only release, while its companion needed nothing further than the President's approval — part of a broader national push on AI governance and regulation.
Ministry of Digital Transition and Administrative Reform — The Rabat-launched 'Morocco AI 2030' roadmap anchors Morocco AI policy, targeting $10 billion in GDP growth, 50,000 AI jobs and 200,000 trained AI graduates by 2030. Branded 'AI Made in Morocco', the national AI strategy founds the JAZARI research-institute network and plans sovereign data centres and cloud infrastructure.
Parliament of the Czech Republic — Czechia's Criminal Code (Act No. 40/2009 Sb.) now covers AI deepfake criminalisation under new Section 191a, punishing non-consensual pornographic deepfakes made or shared without the depicted person's agreement. Offenders face up to two years in prison for a standard breach, rising to five years for aggravated cases.
National Council of the Slovak Republic — Slovakia's Act No. 318/2025 Coll., a Slovak AI regulation, folds the EU Artificial Intelligence Act (Regulation (EU) 2024/1689) into domestic law, amending Act No. 56/2018 Z.z. on product conformity assessment. High-risk AI systems face new national conformity procedures, duties for economic operators, and enforcement mechanisms with penalties. AI system compliance follows.
The EU AI Act's high-risk AI rules for Annex III systems, covering migration, border control and law enforcement, missed their original 2 August 2026 start. The Digital Omnibus on AI, Regulation (EU) 2026/1744, moved them to 2 December 2027, taking effect on 27 July 2026, six days before the original deadline.
Essex Police suspended live facial recognition, citing accuracy and bias risks, once it emerged its match threshold had been calibrated for another vendor's algorithm. The UK Information Commissioner's Office then audited six forces and made 107 recommendations, 93 fully and 14 partly accepted, finding no standard practice of routine bias testing.
Connecticut's Chief State's Attorney imposed a statewide moratorium in 2026 on police using AI tools to draft, write or narrate criminal reports, warning that officers would swear under oath to computer-generated text. The King County, Washington prosecutor's office takes the same stance.
Sala de lo Constitucional, Corte Suprema de Justicia de El Salvador — El Salvador's Constitutional Chamber has set ethical and legal criteria for generative AI in public institutions and AI in the courts, ruling in unconstitutionality case 57-2025, sparked by a claim citizens drafted with AI tools. Everyone has the right to a human judge and a human legal representative, and judicial AI use must face strict legal and ethical controls and compulsory human oversight, it held.
Authority for Infocommunications Technology Industry of Brunei Darussalam — Brunei Darussalam's Authority for Infocommunications Technology Industry has overhauled the Guide on Artificial Intelligence Governance and Ethics into a second edition. Generative AI hazards now enter Brunei AI governance, showcased through Bruneian companies already running the principles-based approach. AITI has separately released a Guide on Responsible Use of Generative AI — Brunei Darussalam AI ethics guidance on wording requests, and doing so ethically.
Sindicatura General de la Nación (SIGEN) — SIGEN's Resolution 197/2026, dated 22 June, introduces formal AI governance in Argentina's public sector via structured implementation controls across every national government body. The AI Controls Guide (Guía de Controles Inteligencia Artificial) applies risk-based oversight to AI systems, tying each to institutional goals, legal compliance, and standards of integrity, transparency and accountability.
Office of the Data Protection Commissioner — AI governance in Kenya now carries a formal framework: the Office of the Data Protection Commissioner (ODPC) has released its Guidance Note on Artificial Intelligence, binding banks, insurers, schools, hospitals, employers, retailers, telecoms operators and government bodies to the Data Protection Act wherever they process personal data through AI. Rather than new rules, it maps AI data protection compliance onto existing duties, effective 4 September.
Conseil d'État Malagasy — Madagascar's Conseil d'État Malagasy and 85 magistrates adopted an Ethical Charter for AI use in courts at a UNESCO seminar on artificial intelligence and the rule of law in Antananarivo. It sets out permitted and forbidden applications for AI in judicial decision-making by magistrates, closing a gap left by prior unguided use, and rests on the magistrates' code of ethics.
Agencia de Transformación Digital y Telecomunicaciones — A Mexican federal push aims to make the state a builder, not a buyer, of AI systems, under Mexico AI policy set out by the Agencia de Transformación Digital y Telecomunicaciones (ATDT) in its Plan Nacional de Inteligencia Artificial. Three pillars anchor it: ethics and law, homegrown talent and public software, and sovereign infrastructure for technological sovereignty. Officials envisage a future federal Artificial Intelligence law demanding openness, explainability, fair and proportionate use, human review and shared accountability — echoing broader national AI strategy debates elsewhere.
High Authority for Audiovisual Communication (HACA) — Ahead of Morocco's September 2026 legislative elections, regulator HACA (the High Council for Audiovisual Communication) issued Decision No. 50-26, barring broadcasters from airing electoral material generated or altered by artificial intelligence that risks misleading citizens. This Morocco AI regulation still allows synthetic material for teaching or explanatory purposes, provided it carries a permanent, unmistakable label flagging its AI origin under the new AI-generated electoral content rules. The measure sits within wider guidance the regulator has set for election-period media coverage.
Ministry for Regulation — New Zealand's Ministry for Regulation now steers regulatory leaders toward safer AI in regulatory functions, via Responsible AI in action: Guidance for Regulators. Version 1.1 adds Māori data sovereignty material, citing Te Kāhui Raraunga on Māori data governance expectations. Aim: confident AI adoption in regulation.
Ministry of Justice of North Macedonia — North Macedonia's Ministry of Justice opened consultation until mid-March on draft changes to the Law on Personal Data Protection, tackling North Macedonia AI regulation for the first time. Filed via the ENER platform in February 2026, the text lets firms cite legitimate interest for machine learning systems personal data processing during training and testing, given safeguards, plus strict data-minimisation caps.
Ministry of Digital Affairs — Taiwan's Ministry of Digital Affairs has drafted an AI Risk Taxonomy and Assessment Framework, an AI risk assessment framework under the AI Basic Act aligning domestic rules with international standards. Submitted to the Executive Yuan, it sets four stages — cataloguing AI use cases, identifying risks, assessing them, responding — guiding authorities toward risk-based regulation, part of wider AI risk governance efforts.
Asamblea Nacional de Panamá — The National Assembly passed Bill 588, an adaptive-governance Panama AI regulation, in third debate, sending it to the Executive for presidential sanction. The AI rules rank systems across four impact levels (low, moderate, high, critical) with matching transparency, oversight and safeguard duties. SENACYT chairs a National Commission for Regulatory Experimentation; high-impact decisions need human oversight, and public entities must embed design-to-deployment safeguards.
Ministry of Information Technology and Telecommunication — Pakistan's draft National Data Governance Policy 2026 gives AI regulation for government agencies its own chapter, now open for public comment through the Ministry of IT and Telecommunication. Systems behind legally significant decisions need explainability, ongoing oversight, human review and openness to the public, each entered into a registry the Pakistan Digital Authority runs. Cabinet approval and gazette notification remain pending, so this AI governance framework is not yet binding.
Ministry of Science and Technology — The Minister of Science and Technology's Decision 2847/QĐ-BKHCN of 23 June requires national public-service Artificial Intelligence platforms to score at least 70 of 100 points for official deployment. This Vietnam AI regulation mandates a domestically developed large language model on local infrastructure. The AI rules bind 14 ministries, 3 ministerial-level agencies and 34 localities.
Ministry of Information Communication Technology, Postal and Courier Services — Zimbabwe now has two new bodies overseeing its Zimbabwe AI policy: the National AI Council and the AI Strategy Implementation Office, both created by the Ministry of ICT, Postal and Courier Services. Their mandate covers the National AI Strategy 2026-2030, with the Council steering strategic direction and the Office running monitoring and stakeholder coordination on the ground. Together they form the backbone of the country's AI strategy governance across every sector.
Agence des Systèmes d'Information et du Numérique (ASIN) — Benin's Agence des Systèmes d'Information et du Numérique (ASIN) has set generative AI public agents rules for civil servants in a June 2026 bulletin, 'Billet juridique sur l usage des outils d intelligence artificielle générative par les agents publics au sein de l administration', covering data protection, digital sovereignty, cybersecurity and vendor contract safeguards. Standards include sovereign data hosting, full encryption and ISO/IEC 27001, 27701 and 42001 compliance. ASIN also recommends mapping AI tools in government administration, moving staff off personal accounts onto secured professional subscriptions, and adopting one binding charter.
Hellenic Data Protection Authority (HDPA) — Greece's HDPA outlawed the Hellenic Police's €4 million Smart Policing scheme under Decision 45/2025, banning AI-enabled surveillance in law enforcement. Scanning faces and fingerprints via handheld field gear, the biometric identification system for policing lacked lawful footing, regulators found, and officers were told to keep it off.
Comisión de Regulación de Comunicaciones (CRC) / DNP / MinTIC / MinCiencias — Colombia now enforces ethics rules governing AI in Colombian public entities. The Commission for Communications Regulation (CRC), with the Presidency, MinTIC, MinCiencias and DNP, produced the Guía Ética para la Implementación, Desarrollo y Uso de Sistemas de Inteligencia Artificial en Entidades Públicas de Colombia, setting values for trustworthy public-sector AI, drawing on OECD, UNESCO, United Nations and Alan Turing Institute standards.
Council of Ministers of the Republic of Belarus — Belarus's Council of Ministers has mandated AI in public administration through Resolution No 793, backing surveillance video analytics powered by artificial intelligence and citizen-facing digital assistants with over 1.08 billion Belarusian roubles through 2030. Three strands carry the work: 'Data Economy', 'Digital State' and 'Digital Sovereignty', part of the 'Digital Belarus' programme effective 1 January 2026.
Presidency of Turkey — A Public Artificial Intelligence Directorate General now operates inside Türkiye's Presidency Cybersecurity Directorate, overseeing government AI applications. Mehmet Fatih Kacır said it will draft public-sector AI systems rules, data governance benchmarks for e-government, and coordinate AI deployment across ministries. The National Technology Directorate's name also changed to reference artificial intelligence.
Ministry of State for Technology and Artificial Intelligence (MITAI) — Kamal Shehadi released Lebanon's first draft framework for Lebanon AI regulation, covering artificial intelligence use nationwide as a national AI framework for coming years. A tiered-risk model shapes coverage of health, finance, education and public services, backed by five years of pilots, capacity-building and joined-up government digital systems.
Ministerio de Telecomunicaciones y de la Sociedad de la Información — MINTEL formalised Ecuador AI policy via Acuerdo Ministerial MINTEL-MINTEL-2025-0030, adopting the Estrategia para el Fomento del Desarrollo y Uso Ético y Responsable de la Inteligencia Artificial en el Ecuador (EFIA) as its binding annex. The accord took force upon publication in Registro Oficial Suplemento No. 206 on 19 January 2026, its sole clause ordering the strategy's issuance. EFIA promotes ethical, responsible AI development nationwide.
Office of the National Defense and Security Council — Myanmar's Office of the National Defense and Security Council formed the Supervisory Committee for Legal Awareness Dissemination Using AI Technology under Notification Order No. 120/2025, overseeing the Legal Assistant AI Model. The Ministry of Legal Affairs and the National Artificial Intelligence Technology Development Project jointly run this AI legal awareness, AI-driven legal education effort, discussed at a Nay Pyi Taw meeting.
Ministry of Communications and Information Technology — Qatar has released its draft National Artificial Intelligence Policy for comment, one of six digital policies under a Qatar AI regulation push for the economy. Five pillars anchor it: infrastructure, data, skills, institutional capacity, and governance. Feedback runs via Sharek, part of a unified national AI strategy.
Ministry of Posts, Telecommunications and Digital Economy — The Ministry of Posts, Telecommunications and Digital Economy approved a 2026-2035 national artificial intelligence roadmap anchoring Guinea AI policy. Phases: institutional, regulatory and technical groundwork, then gradual AI use in public services, then regional AI leadership. The UNDP's Artificial Intelligence Landscape Assessment (AILA) methodology shaped this national AI strategy, a first among French-speaking countries.
Government of Quebec — Quebec's Government has ended a March 2025 ban on public bodies using AI virtual assistants, alongside two new papers on AI regulation public sector rules. One directive sets requirements for generative AI governance, risk management and data protection; a second updates guiding principles for trustworthy use. Together they frame how artificial intelligence deployment should proceed across every provincial public body.
Artificial Intelligence Strategic Headquarters (Cabinet Office) — Japan AI regulation now includes national guidance from the Artificial Intelligence Strategic Headquarters under Article 13 of the AI Act, urging voluntary, proactive safeguards from every party in AI research, development and deployment. The framework lists seven pillars — human-centricity, safety, fairness, privacy, security, transparency, accountability — and requires each ministry to name an AI governance officer. It favours soft rules over binding law, relying on self-driven compliance across the sector.
Cuban National Assembly — An AI ethics commission now sits inside Cuba's General Law of Science, Technology and Innovation, covering artificial intelligence research oversight for the first time. The law further binds sector-specific policy to national development goals. Eight chapters and three complementary norms complete the package.
Ministry of Investments, Regional Development and Informatization of the Slovak Republic (MIRRI) — A national AI strategy roadmap now guides Slovakia: MIRRI has released the AI Vision for Slovakia (Vízia AI pre Slovensko), mapping the country's AI adoption ambitions. Six themes anchor it — computing infrastructure, data sovereignty with administration, economic growth, ethical protection of fundamental rights, skills-building, and European and global cooperation — an AI policy framework preceding a fuller National AI Strategy due in 2026.
Ministero del Lavoro e delle Politiche Sociali — Italy's Ministry of Labour adopted Decree No 180, setting Guidelines for the Implementation of Artificial Intelligence in the World of Work, giving employers a phased roadmap for AI in the workplace. The guidelines cover artificial intelligence in employment through alignment with the EU AI Act and GDPR, addressing worker privacy, non-discrimination and human oversight.
Parliament of Vanuatu — Vanuatu now regulates AI regulation through a permit scheme: the Digital Transformation Act No. 42 of 2025 forces any firm supplying AI-powered tools to Ni-Vanuatu businesses or citizens to register under a new ICT category covering 'artificial intelligence (AI) and AI-related data services'. Parliament enacted the law in November 2025, handing the Department of Communications and Digital Transformation power to draft national AI policy frameworks and conduct compliance inspections. Non-compliant firms risk fines reaching Vt20 million.
38th Guam Legislature — Guam AI regulation now runs through a 12-member Guam Artificial Intelligence Regulatory Task Force, born from Public Law 38-77 once Acting Governor Josh Tenorio put his signature on the Guam Legislature's Bill 64-38 on 16 December 2025. That statute slots Chapter 16 into Title 5 of the Guam Code Annotated and pulls together the Homeland Security director, chief technology officer, University of Guam president, legislative speaker, chief justice, and two figures from private industry to sit on this AI regulatory task force. Inside a two-year window the panel owes drafted AI rules covering both government and commerce, plus progress updates every four months.
Spanish AI Supervisory Agency (AESIA) — Sixteen guides now anchor Spain's EU AI Act compliance push, issued on 16 December 2025 by the Spanish Agency for the Supervision of Artificial Intelligence (AESIA) to steer organisations through obligations for high-risk systems under the AI Regulation. Built from findings inside the Spanish AI sandbox, the set opens with an introduction to the framework, adds thirteen technical documents tackling risk management alongside data governance, transparency and cybersecurity duties, and closes with a checklist. None of this AI Regulation compliance guidance is legally binding, though each document aligns its advice with what regulators expect.
Hungarian Parliament / Government of Hungary — A new fixture in Hungary AI policy: the Hungarian AI Council (Magyar Mesterséges Intelligencia Tanács), formed under Act LXXV of 2025 solely to advise, with no decision power. The AI government commissioner leads it, joining ministries, regulators MNB, NAIH and GVH, and bodies like the Hungarian Academy of Sciences and HUN-REN on Hungarian AI governance.
Ministero del Lavoro e delle Politiche Sociali — Italy's Ministry of Labour has established an Observatory on the Adoption of Artificial Intelligence Systems in the World of Work under Ministerial Decree No 173, aimed at tracking AI workplace adoption nationally. The decree names only the creation of this monitoring body as its purpose, leaving the wider adoption of artificial intelligence systems in workplaces otherwise unaddressed for now.
Conseil d'État (Council of State) — France's Conseil d'État binds its administrative judiciary to seven rules on AI in administrative courts, spanning the Council of State, appeal-level courts and the National Court of Asylum. Judges keep final say over rulings, while French staff face new duties on openness and data protection under this judicial AI governance charter.
Supreme Court of India — Justice P S Narasimha now leads the Supreme Court's revamped Artificial Intelligence Committee, a four-member body guiding AI in Indian courts under Chief Justice Surya Kant. Members are tasked with shaping, launching and overseeing artificial intelligence adoption throughout the country's judicial system.
Government of Portugal — Portugal's Council of Ministers adopted the National AI Agenda (ANIA): 32 initiatives spanning four pillars — data, innovation, skills and ethics. This Portugal AI policy targets Europe's top ten digital leaders by 2030, funding a Public Administration Centre of Excellence, faster Portuguese civil-servant AI training, plus health and industry hubs. This national AI strategy leaves the 2030 timeline unstated.
Congreso de la República de Guatemala — AI deepfake regulation Guatemala takes shape through Iniciativa 6657, the Ley de Protección Digital contra Contenidos Falsificados por Inteligencia Artificial (PRODIGI), filed by Julio Portillo of the CABAL bloc. It sets fines up to Q500 mil and three to six years' imprisonment for deepfake creators, requires an AI-generated content watermarking label plus 48-hour removal of flagged posts, and lets the Superintendencia de Telecomunicaciones fine platforms up to Q2 million. Congress has not passed it.
Cabinet of Ministers of Kyrgyzstan — Kyrgyzstan now has its first nationwide AI safety requirements, set by a Cabinet of Ministers resolution grounded in the Digital Code. High-risk AI systems face safety standards for high-risk AI systems: hazard controls, openness, clear reasoning, oversight, precision, dependability, digital resilience and clean data. Owners must verify safety at every lifecycle stage, start to finish.
Presidencia de Uruguay — Decree 276/025 sets AI sandbox testing environments rules in Uruguay, letting the Council of Ministers oversee pilot data and AI schemes under sandbox pilot projects standards. Fairness, non-discrimination, openness and data protection apply, implementing Articles 74-75 of Law 20.212.
Cabinet of Ministers of the Kyrgyz Republic — Kyrgyzstan's Cabinet of Ministers approved this AI regulation Kyrgyzstan measure, Resolution No. 770, setting binding standards for high-risk AI systems requirements: a danger-assessment method plus rules on transparency, explainability, precision, dependability, digital resilience, data integrity and technical records. Grounded in Articles 193 and 194 of the Digital Code, it covers systems deemed to carry elevated risk. It took effect 15 days after the Digital Code's entry into force, around 21 February 2026.
Australian Government — Canberra now handles AI regulation Australia-wide through existing statutes, scrapping ex-minister Ed Husic's mandatory guardrails and standalone AI act. A $30 million AI Safety Institute will flag gaps under this looser AI governance model. The shift, built into the Government's National AI Plan, favours industry growth over binding rules.
French Government / Ministry of Justice — A newly formed directorate, DPIA, now sits inside the French Ministry of Justice's secretariat-general, tasked with pushing AI in judicial administration forward under Élise Farge Di Maria's leadership. Magistrates and court agents are already trialling an AI assistant built on sovereign hosted infrastructure, handling document analysis, drafting, summarising and translation. Rather than a one-off project, this establishes AI adoption in courts through a permanent internal structure.
Government of Hungary — A fresh institutional structure for enforcing the European Union's AI Act now exists in Hungary, thanks to Act LXXV of 2025 — a landmark piece of Hungarian AI regulation. Oversight duties split between two bodies: the National Accreditation Authority handles notification tasks, while a designated ministry office leads on market surveillance. Backing them is a newly formed advisory group, the Hungarian Artificial Intelligence Council, and violators risk penalties climbing to EUR 35 million or 7% of worldwide annual turnover. This EU AI Act implementation brought most requirements into force.
Ministry of Justice and Human Rights, Government of Togo — Togo AI policy is advancing through a Ministry of Justice and Human Rights blueprint steering public deployment of artificial intelligence. Delegates from ministries, agencies and civil society met at a Lomé event, "Artificial Intelligence and Democratic Governance," debating fair algorithms, sound data, citizens' rights and rules for homegrown research.
Ministry of ICT and National Guidance — Uganda's Ministry of ICT has unveiled its first Uganda AI policy step, standing up a taskforce to steer governance. Specialists must draft guidance touching agriculture, education, health, finance and administration, plus privacy, security, misinformation and bias, aiming toward a national AI governance framework for future regulation.
Government of Canada — Canada's Treasury Board President Shafqat Ali unveiled a first-of-its-kind AI Register, giving Canadians AI transparency in government over how federal agencies use artificial intelligence. Hundreds of entries span automated decision tools and machine-generated policy drafting, letting residents track where departments test or deploy these systems. The launch delivers on a pledge inside the federal AI Strategy for the Public Service 2025-2027, promoting broader visibility into government AI deployment nationwide.
States of Guernsey Committee for Home Affairs — Guernsey wants AI deepfake pornography made a crime, under reforms to the Sexual Offences Law 2020 spanning nine areas of sexual offending. Ministers also propose banning cyberflashing, requesting fabricated images, and holding indecent cartoons of children. Police raised the issue after meeting cases of non-consensual synthetic images current rules failed to cover.
Office of the Data Protection Authority — Anyone in Guernsey handling personal data now has a ten-point checklist for AI in data protection, published by the ODPA. It explains keeping automated systems transparent and accountable, meeting AI governance requirements under the Data Protection (Bailiwick of Guernsey) Law 2017.
Personal Data Protection Authority (KVKK) — Türkiye's KVKK released a 15-question guide on generative AI data protection, explaining how such systems process personal information under Law No. 6698. It covers privacy risks and compliance duties for data controllers, part of an AI data privacy compliance push for organisations using large language models.
Eswatini Communications Commission (ESCCOM) — Broadcasters in Eswatini face an AI deep fakes broadcasting crackdown from their regulator, with ESCCOM demanding ethical, transparent handling of the technology to stop misinformation from wearing down public confidence. ESCCOM CEO Mvilawemphi Dlamini framed the message around responsible broadcasting principles, adding that a forthcoming commission framework aims to shield citizens without smothering AI-driven innovation in the sector.
Government of Seychelles — A new statutory body will steer Seychelles AI policy after Cabinet cleared the NISTI Bill, setting up the National Institute for Science, Technology and Innovation to lead national research and innovation efforts. Lawmakers enacted the Act in December 2025, tasking NISTI with drafting strategies for artificial intelligence adoption, digital transformation and other frontier technologies, plus overseeing tech transfer across the islands. This positions the institute as Seychelles' central authority for coordinating AI regulation and innovation planning.
Integritetsskyddsmyndigheten (IMY) — Sweden's IMY has ended its inquiry into Försäkringskassan's AI welfare system, closing the case after the agency pulled the machine learning-based risk profiling tool it used to flag temporary parental benefit (VAB) claims for added scrutiny. Media coverage prompted IMY to open the investigation in June 2025; Försäkringskassan withdrew the tool soon after and ruled out bringing it back, so IMY judged the danger had passed.
Republic of Kazakhstan — Kazakhstan AI regulation gained legal force when Kassym-Jomart Tokayev signed the Law of the Republic of Kazakhstan 'On Artificial Intelligence'. This AI governance framework obliges systems to operate with transparency, accountability and human oversight, outlaws manipulative tactics and unconsented data gathering, and compels labelling of machine-generated material.
Ministry of Digital Affairs — Poland's Ministry of Digital Affairs revised its Policy for the Development of Artificial Intelligence in Poland until 2030 on 17 November 2025, adding indicators to gauge national AI strategy Poland progress. Łódź and the GZM Metropolis join leading centres from 2027; local government pilots PLLuM, and AI HUB Poland readies rollout. An AI adoption roadmap for key industries completes the update.
Parliament of the Republic of Kazakhstan — The President signed Law No. 230-VIII 'On Artificial Intelligence', Kazakhstan AI regulation effective 18 January 2026. The law sorts AI systems into three risk tiers (minimum, medium, high) and bans social scoring and manipulative AI. Synthetic content needs labels, AI system owners face transparency duties, and the AI rules create the National Artificial Intelligence Platform.
Communications and Information Technology Regulatory Authority (CITRA) — CITRA, Kuwait's Communications and Information Technology Regulatory Authority, is overhauling its Kuwait AI regulation to curb data misuse and unify standards nationwide. Khaled Al-Zamil expects specifics within a year, backed by workshops and consultations shaping fresh oversight tools for the technology across industries.
Subsecretaría de Gobierno Digital, Provincia de Buenos Aires — Buenos Aires province has adopted a binding AI risk regulation, Resolution 9/2025, sorting every automated tool its agencies build, buy or run into one of four bands, unacceptable (banned) through to null, echoing the structure of the EU AI Act. Officials must complete algorithmic impact assessments before deploying anything high-risk, register systems in a new Provincial AI Registry, and flag AI clearly whenever residents interact with it. The move grounds Argentina's provincial public administration in this wider AI governance framework.
Municipalidad de la Ciudad de Mendoza — Mendoza's city council enacted Ordenanza N° 4221/2025, a legal framework for AI in government use spanning municipal administration. Officials must guarantee algorithmic clarity, safeguard citizen data, keep automated tools explainable and preserve human review over machine-made decisions, while a new body, the Programa Municipal de Innovación con IA, takes shape inside the city's Comité Local de Inteligencia Artificial. Mendoza sits among the members of the Coalición de Ciudades por la Inteligencia Artificial en Argentina (CIIAR), which steers municipalities on artificial intelligence policy.
Senado de Puerto Rico — 23-3: that margin let Puerto Rico's Senado pass House Bill 347 (PC 347), fixing the territory's first framework for AI policy in government agencies. PRITS — the Puerto Rico Innovation and Technology Service — wins power under amended Law/Ley 75-2019 to steer algorithmic tools across over 100 central bodies. Agencies need PRITS sign-off via an AI authorisation request before launching any automated decision-making tool; La Fortaleza reviews it next.
Government of Nepal — Nepal's Ministry of Communications and Information Technology has launched a National AI Centre embedding artificial intelligence across government and industry, advancing the Digital Nepal strategy. It will draw private-sector participation, positioning AI adoption in Nepal's public sector alongside AI-driven services for enterprise.
Gouvernement français (Conseil des ministres) — The CNIL would lead market surveillance of high-risk Annex III systems plus prohibited practices under article 24 of the DDADUE bill, which names around fifteen national competent authorities, charged with policing the EU AI Act, after going before the Council of Ministers. In this France AI regulation, Arcom oversees media and transparency obligations while the DGCCRF coordinates as single point of contact. The Senate approved an amended text on 17 February 2026; the National Assembly still has to examine the AI rulebook.
Government of Angola (MINTTICS) — MINTTICS, Angola's Ministry of Telecommunications, Information Technologies and Social Communication, opened public feedback in 2025 on an 86-article draft Artificial Intelligence Law, marking the country's first move toward Angola AI regulation. Companies deploying AI systems would face binding duties around openness, fairness, safety and answerability, while individuals gain a legal right to reject decisions made entirely by machines. Wilful misuse carries penalties of up to 12 years' jail time for people and fines reaching 1.5 billion kwanzas for corporations under this proposed AI governance framework.
Ministry of Electronics and Information Technology (MeitY) — India AI regulation now rests on the India AI Governance Guidelines, built around seven guiding Sutras from the Ministry of Electronics and Information Technology. Six governance pillars and a phased plan with short, medium and long-term steps flesh out the rulebook. Existing statutes, not a fresh AI-specific law, carry forward an adaptive approach for responsible AI adoption.
Government of Sudan — Three new agencies now sit under Sudan's Ministry of Digital Transformation and Communications, following a 5 November 2025 decree from Prime Minister Kamil Idris that set up the Sudanese Data and Artificial Intelligence Authority alongside two sister bodies. Handling this Sudan AI regulation brief means the authority oversees how national data is managed and governed, while pushing government agencies toward using AI for decisions and building public services. It functions as the state's central AI governance authority, though details on the other two agencies are not covered here.
Digital Jersey — A new body chaired by Digital Jersey, the Jersey AI Council, now steers AI adoption in Jersey. Its founders bridge public and private spheres: Government of Jersey, Jersey Financial Services Commission, Jersey Office of the Information Commissioner, and Jersey Finance, together promoting trustworthy AI uptake across the island.
Government of Hungary — Hungary's Resolution 1405/2025 mandates AI training for civil servants via a free course, 'MI Jövünk - Mesterséges Intelligencia alapjai', reaching staff at state-owned firms and state-funded universities alike. By 31 December 2026 the same decree demands AI teaching in schools and universities, plus vocational bodies. New bachelor's, master's and PhD tracks in artificial intelligence, tied to global and European research networks, are also required.
Government of Hungary — Hungary now requires an AI officer in every ministry under its AI governance in government ministries resolution, adopted 4 November 2025. Counties must name an AI-responsible official by 31 March 2026, and universities must assign AI specialists too, extending AI officer appointments in the public sector beyond central government. Civil servants and state-company staff face mandatory AI training, and by December 2026 ministers must fold AI teaching into education curricula.
Hungarian Parliament — Hungary now has Act LXXV of 2025, the national law implementing the European Union's Artificial Intelligence Regulation. This Hungarian AI regulation names the National Accreditation Authority as notifying body, hands market surveillance to the Minister for Enterprise Development, and sets up a Hungarian AI Council plus sector watchdogs, with the central bank covering financial AI systems. An AI regulatory sandbox and compliance procedures complete the EU AI Act implementation Hungary framework, mostly active from 1 December 2025.
Superintendencia de Bancos (SIB) — Guatemala's Superintendencia de Bancos (SIB) has brought in AI regulation in banking through Resolución JM-98-2025, issued 30 October 2025, giving the country its first legal definition of artificial intelligence. Article 22 obliges banks and financial institutions to build governance structures — policies, procedures and human review — around decisions made by automated systems, under the broader Reglamento para la Administración del Riesgo Tecnológico. The Junta Monetaria's rule, replacing the 2021 framework, pulls Guatemala's banking sector toward global norms for AI-related risk management.
National People's Congress — China's Cybersecurity Law now includes a dedicated artificial intelligence clause, folding AI regulation into national security legislation for the first time rather than standalone rules. The NPC Standing Committee approved the change, effective 1 January 2026, favouring AI governance woven into existing legal structures sector by sector over one master statute.
Hrvatski sabor (Croatian Parliament) — Croatia has created a distinct AI system endangerment offence under Criminal Code Article 215a, covering danger to life, limb or property from an AI system, with six months to fifteen years by severity. Article 87(33) now defines 'AI system' per the EU AI Act, embedding AI-linked criminal liability in penal law.
Agenzia delle Entrate — Generative AI tax administration excludes machine drafting at Agenzia delle Entrate, Italy's revenue authority, barring AI-written assessments, refunds and litigation papers. Complaints over 'hallucinated' assessment notices sparked this internal prohibition, publicly confirmed 17 December. Approved systems still assist productivity or risk analysis under this AI-generated tax assessments rule, though staff must verify every result by hand.
National Assembly of the Republic of Slovenia (Državni zbor RS) — Slovenia's National Assembly adopted the Act on Implementation of the EU AI Act (ZIUDHPUI), the country's Slovenia AI regulation for artificial intelligence oversight. It names authorities to supervise AI systems, sets up a National Council for Ethics in AI, and requires regulatory sandboxes, building the nation's AI governance framework.
Državni zbor Republike Slovenije (National Assembly of the Republic of Slovenia) — Slovenia's National Assembly passed ZIUDHPUI, the Act on the Implementation of the EU Regulation on Harmonised Rules on Artificial Intelligence, Slovenia AI regulation naming enforcement bodies: AKOS, the Information Commissioner, the Bank of Slovenia, the Insurance Supervision Agency, the Market Inspectorate, and new notification bodies. It also creates a National Council for Ethics in AI with power over fines and penalties, part of Slovenia's harmonised AI rules oversight. Signed 31 October 2025, in the Official Gazette 6 November 2025, binding 21 November 2025.
Nigerian House of Representatives — HB 2567, the Nigerian Artificial Intelligence Bill, 2025, sits before Nigeria's House of Representatives, building Nigeria AI regulation around ethical, accountable AI development. Backers call it an AI governance framework covering deployment and oversight nationwide. National Assembly members are still debating passage.
Cámara de Senadores de Bolivia / AGETIC — AGETIC gains oversight of Bolivia AI regulation now that the Senate cleared Proyecto de Ley N° 178/2024-2025, the nation's first sweeping AI rulebook. Within 180 days a National AI Strategy must follow, alongside duties on transparency, security and human rights, plus bans on social scoring, warrantless mass facial recognition and subliminal manipulation. Titled "Promoción, Gestión y Uso de la Inteligencia Artificial," the bill awaits the Chamber of Deputies to finalise Bolivia's artificial intelligence governance framework.
President of the Republic of Uzbekistan — Uzbekistan's Presidential Decree UP-189 sets targets for AI adoption in government: 100 AI projects across public administration and the economy by end-2026, $1 billion in foreign AI investment by 2030, and 15 AI laboratories at higher education institutions by end-2026. A new Advisory Council under the Ministry of Digital Technologies will steer AI implementation in the public sector, backed by the ai.gov.uz portal, a non-governmental AI Alliance, and $100 million from the Reconstruction and Development Fund for AI infrastructure and projects.
Cámara de Senadores, Asamblea Legislativa Plurinacional — The Cámara de Senadores passed Proyecto de Ley N° 178/2024-2025, 'Promoción, Gestión y Uso de la Inteligencia Artificial', a Bolivia AI regulation of 31 articles and two transitional provisions. The bill would create the Autoridad de Regulación de la Inteligencia Artificial (ARIA), ban seven practices including subliminal manipulation and mass facial recognition without a court order, and set duties for high-risk systems. AGETIC would gain powers to supervise and sanction AI (IA) use and would draft the Estrategia Nacional de Inteligencia Artificial within 180 days. The proposed AI law still awaits final approval in the Cámara de Diputados.
Government of Chad — Chad's Government has created an AI policy directorate, the Directorate General for Artificial Intelligence (DG IA), under the Ministry of Telecommunications, Digital Economy and Digitalization of the Administration, by presidential decree. Zara Ahmed Sidick heads the body, the country's first of its kind, tasked with drafting national strategy, flagging priority sectors, gauging AI's impact and scouting automation across public services. Chad AI governance had no comparable unit before.
Digitaliseringsstyrelsen (Danish Agency for Digital Government) — Denmark AI regulation now has a six-part guidance set from the Danish Digital Government Agency, spelling out which artificial intelligence uses the EU AI Act bans outright. One document covers the full list of prohibitions; five companion texts drill into single cases: manipulation causing harm, exploiting vulnerable people, scoring citizens socially, reading emotions automatically. Each companion guide works as a variant entry point into the same banned AI methods.
Government of Zimbabwe — Zimbabwe's Cabinet has backed a National Artificial Intelligence Strategy running 2026 to 2030, part of its Zimbabwe AI policy push to move the economy away from resource dependence and toward knowledge-driven growth. Six pillars anchor this AI roadmap: growing homegrown AI skills, building computing infrastructure, spreading AI use through government and business, and an ethics framework drawing on UNESCO recommendations. A new National AI Council will steer direction, while an AI Strategy Implementation Office handles coordination on the ground.
Cyberspace Administration of China and National Development and Reform Commission — Chinese government AI policy now pushes large AI models across ministries, via joint rules from the Cyberspace Administration and National Development and Reform Commission. Scenario-based adoption covers citizen services, governance and decision support, provinces leading and prefecture cities following. A ban on feeding state secrets into non-classified systems pairs with warnings against hollow 'digital formalism', echoing wider AI in public administration goals.
Government of Malta — Malta has named its Information and Data Protection Commissioner (IDPC) the market surveillance authority for high-risk AI systems under the EU AI Act, per Legal Notice 227 of 2025. This Malta AI Act designation gives the watchdog jurisdiction over deployments in policing, frontier control, immigration, courts, elections, biometric ID and emergency response — the categories the AI Act treats as high-risk. A companion clause in the notice lays out when live facial-recognition scanning of crowds may be authorised, forming part of Malta's AI regulatory compliance regime.
Parliament of Malta — Under Legal Notice 226, the Malta Digital Innovation Authority (MDIA) now leads enforcement of the EU AI Act domestically, a shift central to Malta AI regulation. Officials tasked the MDIA with market surveillance, notifications and acting as chief contact for AI Act queries, plus overseeing a sandbox for artificial intelligence trials. Breach penalties also fall within its remit, anchoring this Maltese AI governance framework.
Commission de Protection des Données Personnelles (CDP) — Senegal's CDP publicly denounced AI misuse on social media on 8 October 2025, citing manipulated videos and photos of public figures and ordinary people circulated without consent. Commission de Protection des Données Personnelles officials tied this AI-generated deepfakes and manipulated media trend to opinion-manipulation tactics, plus dangers of stolen identities and misused personal information. They pressed citizens, creators and platforms alike to jointly guard against such harms.
Ministère de l'Économie Numérique — The Democratic Republic of the Congo (DRC) began drafting its first National AI Strategy on 8 October 2025, backed by President Félix Tshisekedi's $1.5 billion Digital Nation 2030 vision. Augustin Kibassa Maliba, Minister of Digital Economy, opened the inter-institutional commission producing this artificial intelligence policy framework alongside a National Digital Plan for 2026-2030.
Asamblea Nacional del Ecuador — Ecuador's Asamblea Nacional passed a Public Procurement Law reform, 80 votes to none opposed, embedding AI for contract oversight against irregularities, with alerts tied to the Comptroller's Office and Prosecutor's Office. Emergency deals and case files must stay traceable, while a new platform cross-checks AI-driven procurement monitoring data against the SRI, Civil Registry and Comptroller.
Swedish Government — A new law and ordinance for Sweden AI regulation has been put forward by the government inquiry AI-utredningen, aiming to fill gaps left by the EU AI Act. Its blueprint hands PTS the lead role in a national market-control system for overseeing AI, while IMY and Finansinspektionen would jointly supervise high-risk applications. Set out in report SOU 2025:101, titled "Anpassningar till AI-förordningen" (Adaptations to the AI Regulation), the plan further recommends a regulatory sandbox operated by PTS and stronger sanction powers for regulators, all due to take effect from 2 August 2026 under this Swedish AI law.
National Cyber Security Agency (NCSA) — Thailand's National Cyber Security Agency published AI security guidelines Thailand-wide, setting recommendations for shielding AI systems from cyberattacks at every lifecycle stage. Called the AI Security Guidelines, the document leans on ISO/IEC 42001:2023, ISO/IEC 23894:2023 and the NIST AI risk management framework while tackling hazards particular to artificial intelligence, among them poisoned training data, evasion attacks and injected prompts. This gives Thai regulators a reference point for AI cybersecurity risk management rather than relying purely on generic infosec practice.
GovTech Agency, Royal Government of Bhutan — Gross National Happiness, not GDP growth, anchors Bhutan's AI national strategy. GovTech Agency, Royal Government of Bhutan, built NAIS 2025 on eight priority areas spanning farming, healthcare, schooling, tourism and government services, backed by seven pillars: infrastructure, governance, skills and funding. This AI policy stays non-binding.
Federal Court of Canada — Canada's Federal Court will withhold automated decision-making tools from judgments until public consultation occurs, per Interim Principles and Guidelines steering judicial AI governance, which covers automated tools in court rulings. Staff may still tap such systems for translation and clerical duties, so long as humans verify every output.
Federal Court of Canada — Canada's Federal Court bars judicial AI decision-making in judgments and orders without prior public consultation, under its Interim Principles and Guidelines on the Court's Use of Artificial Intelligence. Seven principles govern: accountability, rights, non-discrimination, accuracy, transparency, cybersecurity, human oversight. The standard is human-in-the-loop court decision-making, not automation.
State of California — California Governor Gavin Newsom enacted Senate Bill 53 (Transparency in Frontier Artificial Intelligence Act), frontier AI regulation obliging major developers to post a safety framework. These advanced-model rules create a CalCompute consortium for public computing, route critical safety incidents to the Office of Emergency Services and shield whistleblowers exposing health and safety risks.
Government of Ghana — Ghana now has a national blueprint for building AI systems responsibly: the Ghana AI Practitioners' Guide (GAIPG), produced by Heritors Labs alongside BMZ and GIZ. Unveiled in Accra, it lays out ethical steps for every phase of AI development, from design through deployment, and is intended to shape upcoming legislation and Ghana AI policy through practical AI governance guidance.
Ministry of Industry and Trade of the Czech Republic — Czech AI regulation gains a draft bill: Czechia's Ministry of Industry and Trade wrote ten pages covering EU AI Act enforcement, pending inter-ministerial sign-off. Three watchdogs split duties — the Czech National Bank, the Office for Personal Data Protection and the Public Defender of Rights — plus the Czech Telecommunication Office as sole contact.
Cabinet of Mongolia — Mongolia's Cabinet has cleared the National Strategy on Big Data and Artificial Intelligence, a five-year blueprint of 66 measures under four goals to grow a domestic AI ecosystem. Mongolia AI policy now aims to rank the nation among Asia's top ten in artificial intelligence, the Prime Minister confirmed.
Cabinet of Mongolia — Mongolia's Cabinet adopted the "Digital First" policy, requiring digital technology and AI in government to take precedence across every tier of decision-making, policy formulation and service provision. The framework underpins the country's wider digital transformation in public administration agenda, including phasing out paper-based processes entirely.
Ministry of Justice — Türkiye's Ministry of Justice has set up a new AI and Data Analytics Branch Directorate inside its Information Technology General Directorate, part of a wider push on AI in judicial administration. The unit will oversee tools already summarising lengthy expert reports in seconds, easing clerks' workloads under this AI in court systems effort, while judges, prosecutors and lawyers keep final say over rulings. Minister Yılmaz Tunç framed the technology as assistive rather than decision-making, stressing that human legal professionals remain firmly in charge.
Government of Portugal — Anacom, the National Communications Authority, now serves as Portugal AI regulation's designated watchdog and contact for EU AI Act enforcement. Officials revealed the pick at the regulator's yearly gathering, citing its lengthy record overseeing telecom and tech markets. Lisbon finalised this weeks after Brussels' 2 August cutoff had lapsed.
Federal Ministry of Communications, Innovation and Digital Economy — Nigeria AI policy now runs on five pillars set out in the National Artificial Intelligence Strategy 2025-2029: infrastructure, ecosystem growth, sector-wide adoption, ethics safeguards and governance. The Federal Ministry of Communications, Innovation and Digital Economy issued the plan on 19 September 2025 with NITDA as lead implementer. Officials call it a living national AI strategy meant to evolve rather than stay fixed.
Norwegian Digitalisation Agency (Digdir) — Norway's Digitalisation Agency (Digdir) has set up KI Norge, a national hub driving AI adoption public sector activity and private organisations across the Norwegian economy. The initiative pairs shared guidance with a regulatory sandbox run alongside Nkom and Datatilsynet, plus links to industry, universities and researchers, backed by NOK 30 million in government funding covering the hub, sandbox and supervisory function. Full operations for this artificial intelligence deployment in government begin in August 2026.
Italian Parliament — Italy AI regulation now includes a milestone: the Italian Parliament enacted the nation's inaugural comprehensive statute governing artificial intelligence, effective from 10 October 2025 and harmonised with the EU AI Act, positioning Italy ahead of every other EU member state on this front. Two bodies now oversee compliance, the Agency for Digital Italy (AgID) and the National Cybersecurity Agency (ACN), while lawmakers committed to refreshing a national AI framework every two years. Public funding worth one billion euros will back AI innovation alongside cybersecurity work and startup growth.
Council of Ministers, Lebanon — Lebanese ministers backed a bill making the State Ministry a full Ministry of Technology and Artificial Intelligence (MITAI), pending parliament vote. Kamal Shehadi said it will set national AI strategy, build rules, launch one public platform, fight cybercrime — central to Lebanon AI regulation and artificial intelligence governance.
National Privacy Commission (NPC) — Advisory No. 2025-02 binds Philippine controllers handling AI in data privacy to a life-cycle safeguard duty, per the National Privacy Commission's Guidelines on Privacy Engineering in Systems Life Cycle Processes. Coverage runs from early planning through ongoing upkeep. Fines for breaches reach PHP 5 million per instance, tied to broader AI governance and privacy engineering standards.
Saudi Authority for Intellectual Property (SAIP) — Saudi Arabia's AI copyright infringement case ended with a SR9,000 fine, the country's first such enforcement action. The Saudi Authority for Intellectual Property found that a person had altered a private photograph with artificial intelligence, then republished it commercially without the owner's consent — an AI-generated content copyright violation under national law.
Minister for Government Modernisation — Belgium's DeepSeek AI ban takes effect 1 December 2025 federally, ordered by the Minister for Government Modernisation. Belgian civil servants must delete DeepSeek applications and stop running them. Cybersecurity concerns drove the DeepSeek AI system prohibition: a June 2025 review by the Centre for Cybersecurity Belgium found troubling patterns in how the Chinese chatbot handles transmitted data.
Cabinet of Trinidad and Tobago — A new Inter-Ministerial Steering Committee on Cybersecurity and Artificial Intelligence gives Trinidad and Tobago's Cabinet a single body for AI policy coordination. It handles cyber-incident response and keeps laws on AI regulation aligned across government departments.
Presidencia del Consejo de Ministros (Presidency of the Council of Ministers) — A new Peru AI regulation now governs artificial intelligence through Supreme Decree DS 115-2025-PCM, which puts AI Law 31814 into force with tiered rules matched to risk level. National authority status goes to the PCM's Secretariat of Government and Digital Transformation, while the freshly created National Centre for Digital Innovation and AI (CNIDIA) takes on implementation duties. Sectors deemed high-risk must satisfy demands for algorithmic audits, human oversight, transparency and data protection — together forming the wider AI governance framework now taking shape.
Conseil Régional de La Réunion (Région Réunion) — Conseil Régional de La Réunion stood up its first Regional Data and Artificial Intelligence Committee, uniting public bodies with private firms behind regional AI governance built for the common good. A shared framework guides the panel toward open, dependable, ethically handled records, casting AI as an engine of local innovation rather than a mere administrative add-on beyond territorial data governance. This step extends the Regional Digital Strategy already in force since September 2023.
Presidencia del Consejo de Ministros — Decreto Supremo N° 115-2025-PCM implements Ley N° 31814 through six titles and 36 articles. This Peru AI regulation governs AI systems throughout their lifecycle, requiring human oversight for high-risk systems in justice, finance, health, education and basic services. Under these AI rules, every public entity must adopt an institutional AI policy.
Presidential Council on National Artificial Intelligence Strategy — South Korea now runs national AI policy through one presidential body: the Presidential Council on National Artificial Intelligence Strategy (국가인공지능전략위원회), under President Lee Jae-myung. As top AI governance authority, it sets overall strategy, coordinates ministries, and tracks implementation. Its 34 private-sector members split into 8 subcommittees: industry AX, public AX, tech/infrastructure, data, society, global ties, talent, defence/security.
French Government (Ministry of Economy and Finance) — Mayotte's AI policy advisory council, CIANum (Conseil de l'intelligence artificielle et du numérique), weighs how digital tools and AI touch society, the economy and territories. Clara Chappaz founded it 12 June 2025; a 4 September 2025 decree updated the 8 December 2017 text, cementing its legal status. Succeeding the Conseil national du numérique, CIANum gained members by order on 8 September 2025 — Mayotte's AI and digital governance body.
Qatar Cabinet — Qatar's Ministry of Communications and Information Technology gains a new administrative unit after the Cabinet approved a draft resolution creating the National Center for Artificial Intelligence. The move restructures the ministry to define the centre's remit, forming part of a broader push on Qatar AI governance to coordinate AI development and oversight across government entities. Officials frame the body as a coordinating hub for national AI regulation efforts rather than a standalone agency.
Government of Hungary — Government Resolution 1324/2025 sets Hungary AI policy, renewing the AI Strategy 2025-2030 and targeting 15% AI-driven GDP growth by 2030. The update aligns 2020's version with new EU AI Act requirements, adding yearly reviews. The Minister of National Economy and AI Government Commissioner lead the national AI roadmap.
Agencia Nacional de Inteligencia Artificial (ANIA) — ANIA is El Salvador's single AI compliance contact under the Law for the Promotion of Artificial Intelligence and Technologies, with a 12-month grace period sparing developers fines while joining the National Registry. This AI regulation binds automated tools deciding on health, finance, jobs, schooling, housing, transport or public services to warn those affected and allow challenges.
Ministry of the Economy and Innovation (EIMIN) — €3 million in EU funding will back Lithuania's new AI regulatory sandbox, making it one of Europe's earliest adopters of such a scheme, per the Ministry of the Economy and Innovation. Small and medium enterprises gain a supervised space to build and refine AI tools before commercial launch, with the Innovation Agency and Communications Regulatory Authority (RRT) as watchdogs. The initiative reflects a wider Lithuanian AI policy push toward safe AI system development ahead of market entry.
Government of Antigua and Barbuda — Gaston Browne says Antigua and Barbuda has drafted an AI regulation bill to guard against destructive misuse of the technology. Speaking on Point FM, the prime minister linked the draft law to fears that unchecked systems could cause harm, while stressing plans to tap AI gains for creative sectors and government services. The artificial intelligence legislation balances these dangers against economic opportunity.
Austrian Federal Administrative Court (BVwG) — AMS has won a GDPR fight over AI in employment services: Austria's Federal Administrative Court ruled AMAS doesn't breach Article 22's automation ban, since staff kept genuine power to overrule its verdicts. A 2020 Data Protection Authority ban on the automated decision-making system in public employment, in force since January 2021, is now void.
National Diet of Japan — Full enforcement of Japan AI regulation arrived on 1 September for Act No. 53 of 2025, the country's first dedicated statute on artificial intelligence. Under this AI-related technology promotion law, a Cabinet-level AI Strategic Headquarters gets created, an AI Basic Plan becomes mandatory for officials to adopt, and authorities gain power to issue guidance, collect information and probe instances where AI causes rights violations, extending direction to companies and ordinary citizens alike. Businesses face no fines or bans under the legislation, only a request to cooperate with state initiatives, having been promulgated earlier on 4 June 2025.
National Agency for Information Society (AKSHI / NAIS) — Public feedback opened 28 August on Albania's five-pillar National Artificial Intelligence Strategy 2025-2030, built with World Bank backing. It aims to steer infrastructure spending, data rules, skills training and AI in the public sector while matching the EU AI Act, with rollout eyed for energy, procurement, tax-evasion checks, healthcare, farming and schooling, part of a wider public sector AI integration push. Formal approval is still pending.
State Council — The State Council Opinions on Deeply Implementing the AI+ Action (Guo Fa [2025] No. 11) set phased targets to 2027, 2030 and 2035 for China AI policy. This national AI strategy spans six priority areas: global cooperation, governance capacity, public welfare, consumption quality, industrial development and scientific research.
Agencija za zaštitu osobnih podataka (AZOP) — Croatia's AZOP has published FAQs addressing AI data privacy compliance questions under GDPR, covering three practical areas: classifying when AI tools process personal data, assigning controller versus processor responsibility, and triggering mandatory impact assessments. This is applied artificial intelligence governance—existing law explained, not new legislation created.
Instituto Nacional Electoral — Mexico's National Electoral Institute (INE) has become the first public body in the country to formally regulate AI use, adopting a new framework for AI governance in Mexican elections plus a 2025-2026 Institutional Work Programme. The framework, formally titled "Lineamientos y principios para el desarrollo estratégico y uso regulado de la Inteligencia Artificial", sets legality, transparency, ethics and institutional efficiency as the principles guiding how the institute builds, deploys and supervises its AI systems. Its leadership described the move as pioneering AI regulation nationwide.
Ministry of Communications and Innovation, Government of Botswana — The Ministry of Communications and Innovation is finalising a draft National Artificial Intelligence Policy after a UNESCO AI Readiness Assessment endorsed preparedness, Assistant Minister Mr Shawn Nthaile confirmed. Botswana AI policy will steer ethical, inclusive, secure deployment to national development goals. This AI governance text remains unenacted and unpublished.
Oliy Majlis (Parliament of Uzbekistan) — Two-stage passage in Uzbekistan gives artificial intelligence a formal legal footing: deputies cleared the bill through the Legislative Chamber on 12 August 2025, before the Senate signed off on 1 November 2025. Built as an amendment to the Informatization Act, the new Uzbekistan AI regulation defines AI in law, forces labels onto AI-generated content, and blocks any decision touching personal rights from resting on AI alone. Handling someone's data unlawfully through AI now carries administrative penalties, closing a gap the earlier act left open — an AI governance law shaped by parliament rather than by regulators alone.
Government of Nepal — Nepal now has a National Artificial Intelligence Policy 2025, giving the country legal and regulatory scaffolding for AI across all sectors under this new Nepal AI regulation. An AI Regulatory Council led by the communications minister will steer implementation, backed by a National AI Centre under the Ministry of Communication and Information Technology, with workforce skills for AI a stated priority — part of Nepal's wider national AI governance strategy.
Présidence de la République Gabonaise — Gabon's government has enacted Ordonnance n° 0006/PR/2025, signed by President Brice Clotaire Oligui Nguema and published in the Journal Officiel, setting a legal framework for public-sector digitalisation. A dedicated "Smart-Gouvernement" component requires administrative bodies to deploy artificial intelligence and emerging technologies so services become proactive rather than reactive, marking a formal push for AI in government administration and AI-driven public administration more broadly. The decree also lays down principles covering digital inclusion, interoperability, security and data sovereignty.
Parliament of Armenia — Armenia's Ministry of Internal Affairs holds nonstop, 24/7 entry to a facial recognition surveillance grid, thanks to Law on Police amendments parliament cleared in March 2025. That biometric camera grid reaches airports, parking zones, cultural sites, transit hubs and municipal or state buildings nationwide. Civil society groups warn the statute lacks real safeguards and any AI-specific oversight.
Personal Information Protection Commission — Chief Ko Hak-su's Personal Information Protection Commission in South Korea introduced its "Guidelines on Personal Data Processing for Generative AI", targeting AI data privacy compliance across how generative systems handle personal information from concept to deployment. Four sequential phases structure the document: defining purpose, mapping strategy, building and training models, then applying and managing them — each carrying baseline protections and lawful-basis tests for training data. A designated chief privacy officer anchors accountability throughout, reflecting wider generative AI governance goals that weigh innovation against individual data rights.
Office of the Prime Minister, Federal Government of Somalia — Somalia's Office of the Prime Minister issued Regulation LR.183 (Xeer-nidaamiyaha Dhijitalaynta Nidaamka Maareynta Maaliyadda Guud), covering digitalisation of the public financial management system. Chapter 6, Article 34 addresses AI in public financial management across government entities, marking one of Somalia's earliest formal nods to AI governance in government finance in such rules.
Parliament of the Czech Republic — Czechia's amended Act No. 110/2019 Sb. on Personal Data Processing, introduced through Act No. 230/2025 Sb., permits AI biometric identification police use at international airports for security screening in real time. Passed on 11 June 2025, the legislation labels qualifying technology an "isolated system", tying the definition to remote biometric recognition tools described under Article 3 point 42 of the EU AI Act. The measure narrows automated identity-scanning of travellers to airport security contexts rather than broader public surveillance.
Comisión Presidencial de Gobierno Abierto y Electrónico (GAE) — GAE, Guatemala's Comisión Presidencial de Gobierno Abierto y Electrónico, is leading a participatory Guatemala AI strategy process, backed technically by UNDP, to shape the Estrategia Nacional de Inteligencia Artificial (ENIA). Three phases structure it: Phase 1, AI Landscape Assessment (AILA), August-October 2025; Phase 2, co-building vision and priorities, March-May 2026; Phase 3, drafting roadmap, oversight rules and tracking mechanisms, July-October 2026. Guatemala's national AI policy remains unfinished.
Agency for Innovation and Digital Technologies under the President of the Republic of Tajikistan — Tajikistan AI policy now runs through a new Interdepartmental Commission for the Regulation of Artificial Intelligence, formed by the Agency for Innovation and Digital Technologies under the President of the Republic of Tajikistan. It coordinates digital technology rollout and drives the AI regulation strategy set out to 2040.
Police of the Czech Republic — New judicial sign-off rules under the EU AI Act forced Czech officers to halt facial recognition airport surveillance at Prague's Václav Havel Airport. Running since 2019, the biometric setup had caught 155 people wanted on arrest warrants before shutdown.
Personal Data Protection Bureau (PDPB) — Nine Asia-Pacific privacy regulators, led by Macao's Personal Data Protection Bureau (PDPB) and Hong Kong's Privacy Commissioner for Personal Data (PCPD), now back a shared framework for AI data anonymisation. Their "Guide to Getting Started with Anonymisation" pushes organisations to scrub identifying details from records before these enter AI models, walking through five phases: locating the relevant data, wiping direct identifiers, deploying anonymisation tools, gauging re-identification exposure, and taming that exposure over time. Put simply, anonymising personal data for AI model training turns into a mandatory groundwork step.
Jogorku Kenesh (Parliament) and President of the Kyrgyz Republic — Kyrgyzstan AI regulation now runs through Digital Code No. 178, with Chapter 23 covering Articles 191 to 197 under Artificial Intelligence Systems. Parliament passed it 18 June 2025, President Sadyr Japarov signed it 31 July 2025, and it took force 6 February 2026. The chapter sets an AI systems risk classification, tougher duties for high-risk categories, and disclosure requirements for AI use.
Ministry of Justice — England and Wales will see AI in justice system tools launched under a new Justice AI Unit and Action Plan spanning courts, tribunals, prisons and probation, the Ministry of Justice confirms. Uses include staff transcription, prison capacity scheduling and public chat assistants. Ministers call the shift toward algorithmic courtroom support measured, not hasty.
Government of Pakistan — Pakistan's cabinet has unanimously cleared the National AI Policy 2025, a Pakistan AI policy meant to grow a full artificial intelligence ecosystem nationwide. It targets one million skilled AI professionals by 2030, 1,000 homegrown AI products within five years, 50,000 civic AI projects, and 3,000 yearly scholarships, steered by a new AI Council under a Master Plan and Action Matrix, plus dedicated Innovation and Venture Funds.
Congress of Colombia — Colombia's Congress toughened penalties for AI-enabled identity theft under Law 2502, amending Article 296 of the Criminal Code so artificial intelligence use turns identity fraud into an aggravating offence. The statute defines deepfakes and gives the government two years to draft an AI regulation policy against such fraud, covering detection tools for the Prosecutor's Office and Police.
Azerbaijan Standardisation Institute (AZSTAND) — Azerbaijan now has a national AI management system standard. AZSTAND adopted AZS ISO/IEC 42001:2025, "Information Technology – Artificial Intelligence – Management System," after review by committee AZSTAND/TK 05. It covers AI risk, ethics, transparency and a system's full lifecycle, and now sits in the State Fund of Normative Documents as the country's official AI governance benchmark.
Information and Communication Technology Division (ICT Division) — Bangladesh set up the National AI Policy Committee to steer its Bangladesh AI policy work. The panel mixes officials, academics, industry figures and international experts, and has run three steering sessions: the 1st, 2nd and 3rd meetings, in August and October. These sittings mark early momentum on national AI regulation.
Information & eGovernment Authority (iGA) — A four-pillar structure — legal compliance, adoption, public awareness, cross-border cooperation — now binds every government body under this Bahrain AI policy. Bahrain's Information & eGovernment Authority enforced it with the GCC's Guiding Manual on the Ethics of Artificial Intelligence Use, linking national AI regulation to Economic Vision 2030 and the UN Sustainable Development Goals.
Government of Algeria — Algeria's Law No. 25-11 narrows the automated-decision rule in the country's data protection framework, Loi 18-07, under Algeria AI regulation: the blanket ban on decisions made solely by automated processing now covers only criminal-justice bodies. It adds a formal definition of 'profilage' and requires a Data Protection Officer, tightening algorithmic decision-making oversight in criminal justice while easing it elsewhere.
Grand National Assembly of Turkey (TBMM) — Turkey's Grand National Assembly received the country's first comprehensive AI regulation Turkey has proposed, on 23 July 2025: the Bill of Law on Amendments to the Turkish Penal Code and Certain Laws. This AI governance legislation gives artificial intelligence a formal legal definition, splits criminal liability between developers and users, and forces deepfake material to be labelled within six hours. The Information and Communication Technologies Authority gains power to fine offenders up to 10 million lira; the bill remains under committee review.
The White House — "Winning the Race: America's AI Action Plan" spans three pillars — innovation, AI infrastructure, and diplomacy and security — via over 90 federal policy actions. A January order prompted this White House blueprint, now the United States' national US AI policy, replacing scattered US AI regulation with one roadmap.
CNIL (French data protection authority) — France's CNIL, the French data protection authority, wrapped up guidance on AI regulation in France, explaining to developers how GDPR obligations bear on building models. Emerging from a GDPR compliance for AI systems consultation open to the public, the guidance tackles cutting down data gathered, narrowing its purpose, fixing retention periods, safeguarding individual rights, and grounding web scraping in legitimate interest. CNIL flagged that industry-specific rules would arrive afterward, in the second half of 2025.
Ministry of Digital Governance — The Ministry of Digital Governance houses a Special Secretariat for Artificial Intelligence and Data Governance (Joint Ministerial Decision 1961/2025, Government Gazette B' 3804). Its job: shaping, coordinating and delivering Greece AI policy plus public data governance. Via Directorates for AI and Data Governance, it voices national AI policymaking on the EU AI Board and in international AI fora.
Ministry of Digital Affairs — Taiwan's Ministry of Digital Affairs opened consultation on the draft Act on Promoting the Development of Data Innovation and Utilization, gathering comment through 15 August 2025 for a framework tackling AI innovation policy. Agencies would gain Chief Data Officers and an Executive Yuan Data Innovation and Utilization Advisory Council on data innovation and use, plus firmer legal footing for data-sharing.
Belgian Federal Government — Belgium's federal government now runs on a shared ethics code: forty public organisations back the new Charter for Responsible AI Use in Government Services, a project Minister Vanessa Matz put forward. Fairness, transparency and human oversight anchor the rules guiding artificial intelligence rollout across federal government AI. The framework aims to steer how public sector artificial intelligence gets used nationwide.
Tribunal Supremo Popular (Supreme Popular Court of Cuba) — Cuba's Supreme Popular Court now has judicial guidance on AI in judicial justice, penned by judge Yojanier Sierra Infante. A single limit governs everything: artificial intelligence stays a supporting aid within the court system, forbidden from standing in for the magistrate or judge's own verdict.
Czech government — Chinese law forces DeepSeek to cooperate with state authorities there, exposing user data to unauthorised access, so Czechia's National Cyber and Information Security Agency (NÚKIB) pushed for an outright block. That advice became a nationwide directive restricting AI use in government agencies, shutting the Chinese firm's tools out of every ministry, an AI ban in public sector systems targeting one named vendor rather than AI broadly.
Bank of Russia — Russian financial institutions now have Bank of Russia guidance on AI in finance: a Code of Ethics built on five principles — human-centricity; fairness; transparency; security; and, separately, responsible risk management. Firms should disclose AI use to clients, allow opt-outs, audit quality, protect data and manage risk. These AI systems in the financial market guidelines remain non-binding.
Føroya Landsstýri (Government of the Faroe Islands) — Ruth Vang's Ministry of Finance now owns AI policy area responsibility in the Faroe Islands, folding it into the existing IT brief. Regulations confirm the shift, placing artificial intelligence governance among the ministry's standing duties rather than treating it as a passing research topic.
New Zealand Government — New Zealand's Government unveiled 'Investing with Confidence', its national AI strategy, branding the nation a 'sophisticated adopter' of machine intelligence rather than an architect of foundational models. Existing statutes and a light-touch, principles-based stance replace any push for fresh AI-specific rules. Science and Technology Minister Shane Reti flagged 8 July 2025 as the latest such rollout among OECD members, leaving New Zealand's AI policy framework trailing every peer nation.
Ministry of Business, Innovation and Employment (MBIE) — MBIE's Responsible Artificial Intelligence Guidance for Businesses gives firms a starter kit for AI adoption in business: risk toolkits, checklists, lifecycle controls. Tied to New Zealand's AI strategy, the non-binding framework favours light-touch, principles-based oversight, steering artificial intelligence for companies toward safer, ethical use.
Ministry of Posts and Telecommunications (MINPOSTEL) — Yaoundé hosted the unveiling of Cameroon's national artificial intelligence strategy, SNIA, at the National Consultations on AI (CONIA 2025), led by Minister Minette Libom Li Likeng of the Ministry of Posts and Telecommunications. Seven pillars structure this AI policy roadmap: sovereignty and governance (proposing a Cameroonian AI Authority and Presidential Council on AI), digital infrastructure and data, multilingual inclusion (featuring GPT Cameroon), sovereign technology systems, workforce and research, sector innovation, and cross-border partnerships. By 2040 officials want 60,000 trained specialists, 12,000 new jobs, and 0.8 to 1.2 percent GDP growth.
Vláda Slovenskej republiky — On 4 July 2025 Slovakia named Radoslav Štefánek as its first Government Plenipotentiary for Artificial Intelligence, steering Slovakia AI policy across ministries, regional bodies, companies and universities while representing the country in international AI forums. A cabinet resolution created the post, run as an advisory arm inside MIRRI, where Štefánek already holds the State Secretary title. This reflects a national AI coordination model rather than a standalone ministry.
Bundesnetzagentur (BNetzA) — Germany's Federal Network Agency (Bundesnetzagentur) now runs a KI-Service Desk helping firms and public bodies through EU AI Act compliance. Its Compliance Compass tool flags whether an AI system sits within regulatory scope, faces transparency rules, or ranks as high-risk, part of wider AI regulation guidance nationally. Free AI literacy courses, required under Article 4 since 2 February 2025, complete the service.
Government of Iceland (Ministry of Culture, Innovation and Higher Education) — 20 fully financed measures anchor Iceland's Action Plan, an Iceland AI policy grouped into five priority areas: societal benefit, business competitiveness, education, reimagined public services, and healthcare. Rolled out in 2025 as a national AI strategy, it spans two years, undergoes six-monthly checks, and faces full assessment by mid-2027. Logi Einarsson framed it as vital for a small, technologically advanced nation pursuing its potential.
Ministry of Post and Telecommunications (MPTC) — The Ministry of Post and Telecommunications opened its draft National Artificial Intelligence Strategy 2025–2030 to public consultation in June 2025; over 300 participants refined it at a national workshop. This Cambodia AI policy has six priorities: ethical and responsible AI, human resource development, sectoral adoption, digital government, data and infrastructure, and collaboration, research and innovation. The national AI plan awaits approval from the Digital Government Committee and National Digital Economy and Society Council.
Conseil des ministres de la République du Congo — Congo's Council of Ministers approved statutes founding a new AI research centre, the African Centre for Research in Artificial Intelligence (CARIA), as a public administrative body. The decree fixes how this artificial intelligence research institute will function, be steered and be supervised.
Government of Norway — Norway's Ministry of Digitalisation and Public Governance launched a three-month public consultation on Norway AI regulation, transposing the EU AI Act into a draft domestic statute. Officials hope KI-loven, the Norwegian AI Act, takes effect by summer 2026, aligning with the EU AI Regulation's main rollout.
Ministry of Justice and Public Security — Brazil now has dedicated rules for AI in criminal investigations under Portaria 961, issued 30 June 2025 by the Ministry of Justice and Public Security. Police at federal, state and municipal levels gain authority to apply such tools, though real-time facial-recognition scanning of the public is banned unless officers are hunting a missing person, protecting a crime victim, or facing an imminent threat. Pulling confidential records through automated systems for confidential records still demands a judge's sign-off.
National Assembly of Kenya — Kenya AI regulation moved ahead when the National Assembly told the Ministry of Information, Communications and the Digital Economy to build ethical standards for artificial intelligence, plus a public education push. MP Marianne Kitany's proposal, aired 18 June 2025, won approval weeks later amid warnings over disinformation, algorithmic bias, privacy violations and vanishing jobs — concerns behind the AI regulatory framework.
Council of Ministers of Bulgaria — EU AI Act Bulgaria implementation now assigns fundamental rights protection to seven named authorities, following Decision No 398, adopted by the Council of Ministers. Oversight duties spread across the Ombudsman, the Central Election Commission, the Commission for Protection from Discrimination, the Commission for Personal Data Protection, the Commission for Consumer Protection, the State Agency for Child Protection, and the General Labour Inspectorate Executive Agency. Each watchdog now enforces AI Act compliance within its existing legal remit, marking a concrete step toward meeting these obligations nationally.
Agencia de Gobierno Electrónico y Sociedad de la Información y del Conocimiento (Agesic) — Uruguay AI policy takes shape as Agesic drafts legislation governing artificial intelligence use. The proposed law sets boundaries on permissible AI applications, prioritising citizens' rights protections, Daniel Mordecki told the Senate's Science, Innovation and Technology Committee. He called the effort consensus-driven, balancing rights safeguards against opportunities, reflecting AI governance in Uruguay.
Dátueftirlitið (Faroese Data Protection Authority) — Handling personal data through ChatGPT breaches the Data Protection Act, Dátueftirlitið, the Faroese Data Protection Authority, cautioned, since nobody can confirm how such tools retain or secure information. This is core to sound AI in public-sector case processing, it said, urging public bodies to set their own rules before adopting generative AI in government data processing.
Ministry of Communication and Information Technology — Tanzania's Ministry of Communication and Information Technology has released a draft Tanzania National Artificial Intelligence Strategy Framework for AI policy in Tanzania, anchored on transparency, fairness and inclusivity. It frames the plan as a route to sustainable growth and sharper public service delivery. The national AI strategy still awaits sign-off.
Ministère de l'Économie, des Finances et de la Souveraineté industrielle et numérique — A decree signed 4 September 2025, revising the earlier 8 December 2017 text, gave France's Conseil national de l'IA et du Numérique (CIAN) its legal footing on AI policy France pursues, replacing the Conseil national du numérique as the country's artificial intelligence advisory body. Clara Chappaz, minister delegate, leads the panel, tasked with examining how digitalisation affects society, the economy and territories such as French Guiana.
Isle of Man Information Commissioner — Six years had passed since the last review from the Isle of Man Information Commissioner's Office when its 2025 report named artificial intelligence and high-risk personal data handling as a growing focus under AI data protection regulation. Government bodies alone accounted for all four Data Protection Impact Assessments submitted that year, with not a single filing from private firms, a pattern the regulator took as evidence that businesses remain unclear on their legal responsibilities toward automated systems. Commissioner Alexandra Delaney-Bhattacharya pressed organisations to grasp that any AI tool processing personal data is subject to Manx law, reinforcing calls for stronger AI regulation compliance island-wide.
Government of the Russian Federation — A new Centre for AI Development now sits under Russia's federal government, formed by Resolution No. 861 atop the existing Analytical Centre — the latest move in Russia AI regulation. Deputy Prime Minister Dmitry Grigorenko leads it, with duties covering AI governance in Russia: picking and scaling standard tools, linking ministries, regions and firms, giving methodological guidance, and handling international cooperation.
Ministry of Digital Economy and Society (MDES) / Electronic Transactions Development Agency (ETDA) — Under Thailand AI regulation plans, ETDA and MDES have floated Draft Principles of the Law on Artificial Intelligence, opened for public consultation with a 2 May 2025 session and a comment window ending a month later. Systems get sorted by risk tier, separating prohibited uses from high-risk categories, while providers must build risk processes matching ISO/IEC 42001, name local representatives, and disclose serious incidents. An AI Act Thailand hearing stage still lies ahead before enactment.
Coordinating Ministry for Human Development and Culture (Kemenko PMK) — Indonesia's PMK Coordinating Ministry launched the Gugus Tugas AI Task Force under Minister Pratikno to speed nationwide AI talent development in Indonesia. Rather than delivering programmes itself, it aligns artificial intelligence strategy across ministries spanning health, learning, heritage, family welfare and faith affairs.
Instituto Nacional de Defensa de la Competencia y de la Protección de la Propiedad Intelectual (Indecopi) — Peru's Indecopi bound itself to human oversight of every algorithm it builds, adopting new rules on AI regulation in public agencies. Resolución N.° 000062-2025-GEG/INDECOPI enacted the Lineamientos para el Uso Ético de la Inteligencia Artificial, setting transparency, security, fairness, privacy and accountability duties across its AI governance.
Government of Mozambique — Mozambique's government is forming a National Artificial Intelligence Commission (CNIA) to handle AI regulation Mozambique-wide and to push adoption across public services. Salim Valá unveiled the commission at Maputo's 12th MOZTECH technology fair, calling it a cross-sector AI governance platform. Decree No. 14/2026 gave it formal standing in April 2026.
Government of the Czech Republic — Under Resolution No. 393, Czechia runs AI governance Czechia via a Government Commissioner for Artificial Intelligence, linking ministries, public bodies and the AI ecosystem. Duties span the NAIS Action Plan, matching rules to the AI Act, and building bodies for artificial intelligence oversight. Jan Kavalírek was first appointed; Lukáš Kačena took over on 12 January 2026.
Government of the Philippines — A formal AI policy Philippines framework is now in place: President Ferdinand Marcos Jr. has signed the National AI Strategy for the Philippines (NAIS-PH) into effect. Unveiled at a sectoral meeting in Malacañang, the blueprint rests on five pillars — infrastructure, workforce development, innovation, data governance and policy, and deployment of AI systems — aimed at turning the country into a globally competitive, inclusive innovation hub by 2028. Better governance and industrial growth both feature as stated goals of the roadmap.
Government of Luxembourg — A Luxembourg AI policy sits inside "Accelerating Digital Sovereignty 2030", alongside linked data and quantum plans. Funding for 2025-2030 targets hands-on trials and real-world use across industries. The Luxembourg AI strategy also promises swift, pragmatic EU AI Act rollout via regulatory sandboxes.
President of the Republic of Kazakhstan — Kazakhstan AI policy now runs through a new presidential advisory body, the Council for the Development of Artificial Intelligence, created by Decree No. 881 and chaired personally by President Kassym-Jomart Tokayev. Seats around the table go to the Prime Minister, the Minister of Artificial Intelligence and Digital Development, and outside foreign experts brought in from beyond government ranks. The panel's remit stretches across drafting national AI strategy proposals, sharpening AI-related legislation, and scrutinising AI projects through expert vetting.
Islamic Consultative Assembly (Iranian Parliament) — Lawmakers in Iran's Islamic Consultative Assembly backed the National Artificial Intelligence Plan 187 votes to 33, one member abstaining, under an Iran AI policy that sets a 2032 deadline for ranking among the top 10 AI-harnessing nations worldwide. The outline requires AI use across 45% of industries, seeks a 12% GDP share from AI, and earmarks roughly $115 million for research and development funding. This marks approval of the plan's general framework, not a completed law, reflecting Tehran's wider push on national AI strategy.
Direction Générale des Douanes — Madagascar now has CEDIA, the Centre d'Excellence des Données et de l'Intelligence Artificielle, its first unit for AI in customs administration. It will oversee governance rules, drive uptake nationwide, and craft tools like Agent SVO, a generative helper for AI-driven fraud detection on questionable receipts.
Folketinget (Danish Parliament) — Denmark AI regulation now runs through Law No. 467, the Act on Supplementary Provisions to the Regulation on Artificial Intelligence, passed by the Folketing and active from 2 August 2025. Three bodies enforce it: Digitaliseringsstyrelsen (Danish Agency for Digital Government) as lead contact, Datatilsynet (Danish Data Protection Agency) and Domstolsstyrelsen (Danish Court Administration), which inspect firms, demand records and levy fines (bødeforelæg) for breaching the EU AI Act — Denmark's version of EU AI Act enforcement.
Folketinget (Danish Parliament) — Denmark AI regulation now sits with three bodies holding fresh powers to probe firms and levy fines. Law no. 467 hands the Danish Agency for Digital Government, the Danish Data Protection Authority and the Danish Court Administration authority over the EU Artificial Intelligence Act, setting AI Act compliance duties from 2 August 2025.
Council of Ministers of Iran — Topping the charts for artificial intelligence capability throughout South-West Asia, while breaking into the worldwide top ten, sits at the heart of Iran's newly cleared Government Document on the Development and Application of Artificial Intelligence Technology — approved by the Council of Ministers on 24 Ordibehesht 1404 (14 May 2025) as the backbone of Iran AI policy. Governance now shifts to a purpose-built coordination body under this AI development and application framework, tasked with folding AI into strategic technologies, tightening data quality and safeguards, scaling up computing resources, lifting technical know-how nationwide, championing knowledge-based ventures, and fostering trustworthy AI adoption. Separate rules address who may reach public datasets, how AI marketplaces function, sandbox testing regimes built for AI, and the credentials demanded of contractors.
Cabinet of Ministers of Ukraine — Ukraine's Cabinet of Ministers adopted the Action Plan for Implementing the Concept for the Development of Artificial Intelligence in Ukraine for 2025-2026, a binding roadmap for Ukraine AI regulation. It covers drafting an AI law, cybersecurity guidance, higher-education standards, procurement analysis tools, defence uses, and aligning domestic rules with international AI policy in Ukraine frameworks. The plan is mandatory for relevant ministries and agencies.
Government of Finland — Government Proposal 46/2025 now sits before Finland's Parliament, establishing an Act on the Supervision of Certain AI Systems that transposes the EU AI Act into domestic law — the core of this Finland AI regulation. Rather than one central regulator, several market surveillance bodies split enforcement duties, with the Finnish Transport and Communications Agency, Traficom, designated as sole point of contact for Brussels. A dedicated Sanctions Board handles penalties topping €100,000, leaving lesser fines to separate authorities under the wider AI Act compliance framework.
Department of Public Expenditure — Ireland's Department of Public Expenditure launched a framework covering AI in public service: the "Guidelines for the Responsible Use of Artificial Intelligence in the Public Service". Jack Chambers presented it with a Decision Framework, a Responsible AI Canvas Tool, and free staff training. This reflects Ireland's push for responsible AI adoption in government, favouring guidance over mandates.
Federal Data Protection and Information Commissioner (FDPIC) — Existing law already covers AI in Switzerland, the Federal Data Protection and Information Commissioner (FDPIC) has ruled: the Federal Act on Data Protection (FADP) applies directly to AI-supported data processing, no new statute required. Under this Swiss AI regulation, anyone building, offering or deploying an AI system must explain what it does, why, and where its data comes from, and people keep the right to know they're talking to a machine, not a person. High-risk processing can be justified with adequate safeguards, but the same data protection AI compliance regime rules out real-time facial recognition and social scoring outright. FADP has been in force since September 2023.
Ministry of Communications and Information Technology — Qatar AI policy now includes a formal ethics framework, released 8 May 2025 by the Ministry of Communications and Information Technology under the title Principles and Guidelines for Ethical Development and Deployment of Artificial Intelligence. Rather than binding rules, the document offers eight guiding values: preventing harm, ensuring robustness, maintaining security and safety, reducing bias, protecting the environment, safeguarding privacy, promoting transparency, keeping humans central to design, and holding humans accountable. Practical advice follows on assessing risks and their impacts, governing data, overseeing systems by humans, and watching over their full lifecycle. Because these AI ethics guidelines carry no legal force, compliance remains voluntary for developers and deployers alike.
Senate of Paraguay — Senator Basilio Núñez wants the Ministry of Information and Communication Technologies (MITIC) enforcing fresh Paraguay AI regulation under a bill that would also launch a National Registry of AI Systems. Fairness, openness and privacy guide its draft code, grading machine-learning tools by hazard tier along lines the European Union already follows. Lawmakers keep reviewing the text.
Government of Colombia (MinTIC/MinCiencias) — MinTIC and MinCiencias filed an AI regulation framework with Congress: risk tiers for AI systems, plus a National AI Authority inside MinCiencias. Julián Molina said it keeps machines within "parameters for human dignity," leaving room for innovation, built on OECD, UNESCO and EU governance standards.
Government of Bangladesh — Bangladesh is first in South Asia to punish AI-powered offences, addressing artificial intelligence misuse under the Cyber Security Ordinance 2025 gazetted 21 May 2025. The interim government's law scraps nine sections of the 2023 Cyber Security Act, grants internet access as a civic right, bans online gambling, and penalises abuse of women and children online.
Løgtingið (Faroese Parliament) — The Faroese Parliament enacted Law No. 78, anchoring AI in public sector services within Digital Faroe Islands 2.0, a 2025-2027 blueprint. Building atop the earlier Talgildu Føroyar scheme, the Ministry of Finance's rules task the Vangin portal with answering residents' private questions via AI-driven public administration tools drawing on official records.
Government of Croatia (Ministry of Justice, Public Administration and Digital Transformation) — The government is drafting a National Plan for the Development of Artificial Intelligence to 2032, plus a 2026 to 2028 Action Plan. Three pillars frame this Croatia AI policy: public policy, investment, conditions for effective AI adoption—education, economic digitalisation, research and innovation support, citizen data protection. Experts drafting the national AI policy first met on 27 May 2025.
Supreme Court of Guam — A hallucination dispute in May 2026 traced back to Administrative Order 2025-001, the rule Chief Justice Katherine Maraman of Guam's Supreme Court had invoked against an attorney over a flawed legal brief. That order, forming the Judiciary of Guam Policy on Artificial Intelligence, binds AI use in court filings to the Guam Rules of Professional Conduct already covering lawyers, jurists and unrepresented parties. Anyone submitting AI-generated legal content in court must personally confirm its accuracy first.
Israeli Privacy Protection Authority — A first-of-its-kind AI regulation Israel measure has arrived from the Privacy Protection Authority, applying the Privacy Protection Law to artificial intelligence systems. Draft rules demand lawful grounds throughout development and use, mandate consent and opt-in steps, let people challenge flawed algorithmic outputs, and ban harvesting personal data from the web for training — forming Israel's Israeli AI data protection guidance. Enforcement under Amendment 13 follows once comments close on 5 June 2025.
Commission for Personal Data Protection (CPDP) — Bulgaria's CPDP has barred shops from using facial recognition AI in retail stores to identify customers, ruling that no GDPR exemption, including theft prevention, permits processing biometric data this way. Because the technology relies on AI, the AI Act also applies, classifying such remote biometric identification technology in shops as high-risk.
Data Protection Commissioner (Komisioneri për të Drejtën e Informimit dhe Mbrojtjen e të Dhënave Personale) — The Data Protection Commissioner's Instruction No. 3 on the Processing of Personal Data through Video Surveillance Systems binds state and commercial CCTV operators, placing facial recognition and video analytics under Albania AI regulation. These AI surveillance rules keep cameras out of bathrooms and changing rooms, forbid live-streaming identifiable footage, delete recordings after 72 hours (extendable to 30 days), and demand notices plus data protection impact assessments.
Ålands landskapsregering — The Government of Åland has adopted an AI policy for public administration, setting rules on permitted AI uses in its agencies. Deployments need a prior risk assessment, and invoice processing is named as a priority use case for pattern recognition cutting manual work. This is Åland's first formal AI governance policy for government operations.
Congreso de la República del Perú — Peru enacted Ley N° 32314 on 29 April 2025, folding AI use into the Código Penal (Decree 635) and the Ley de Delitos Informáticos (Law 30096) as grounds for tougher sentencing. This Peru AI regulation lets judges add up to a third onto prison terms whenever deepfakes or comparable tools drive crimes like fraud, defamation, child pornography or copyright breaches. The reform marks Peru's first statutory move treating AI regulation as a sentencing multiplier rather than a separate offence category.
Cabinet of Zambia — Zambia's cabinet has greenlit, in principle, a bill scrapping the Data Protection Act No. 3 of 2021, swapping it for wider Zambia AI governance legislation spanning machine learning and data analytics. Coverage would stretch past personal information alone, pulling in industrial, public and other non-personal data types under one statute. Ministers cleared this shift toward AI regulation in Zambia during the 8th Cabinet Meeting.
Dutch government (Ministry of the Interior and Kingdom Relations) — Dutch civil servants can now use generative AI in public sector work more freely under new government-wide rules covering municipalities, provinces, water authorities and executive bodies. The policy requires risk analyses beforehand and favours European or open-source tools over other options. This shift toward broader public sector AI adoption gives officials clearer conditions for deploying such tools in daily work.
Senado de la República Dominicana — A Dominican Senate panel is scrutinising Senator Omar Fernández's AI regulation bill, the Organic Law to Regulate AI Systems. The Education, Science Commission is checking how the draft steers AI's rollout nationwide while shielding rights, openness, safety and ethics. Backers claim the AI systems regulation still permits innovation.
Agència Andorrana de Protecció de Dades — A formal probe by the Agència Andorrana de Protecció de Dades, Andorra's Data Protection Agency (APDA), targets Meritxell, the government's AI assistant answering citizen questions on the EU association agreement, examining whether personal data was properly safeguarded in this government AI assistant case. Officials, through spokesperson Guillem Casal, denied storing any personal information and called the scrutiny unexpected, though they admitted an earlier warning had flagged a privacy-policy gap on the official portal, now resolved. At stake: AI data compliance standards for public-sector chatbots.
Digital Policy Office — A move on Hong Kong AI regulation arrived 15 April 2025: the Digital Policy Office issued the Hong Kong Generative Artificial Intelligence Technical & Application Guideline for developers, vendors and users. It draws deployment limits, flags pitfalls like data leakage, model bias and system errors, and sets principles to manage them. Non-binding, it serves as a template for responsible AI adoption in Hong Kong.
Communications, Space and Technology Commission (CST) — Saudi Arabia's Communications, Space and Technology Commission has opened consultation on a draft Global AI Hub Law, an AI data hub regulation closing 14 May 2025. Three hub types, private, extended and virtual, join a data embassy scheme letting foreign entities hold data under home-country law inside Kingdom facilities. This AI infrastructure law targets global status in computing and digital sovereignty.
Senado de Puerto Rico — Puerto Rico AI regulation took a step forward when the Senado passed Senate Bill 68, giving rise to the Ley de Inteligencia Artificial del Gobierno de Puerto Rico. This government AI policy bars agencies from deploying any AI tool that fails ethical benchmarks once 1 January 2026 arrives, with compliance checked through recurring biennial audits. A fresh post, Oficial de Inteligencia Artificial, sits inside the Puerto Rico Innovation and Technology Service (PRITS) alongside a new AI Advisory Committee, and PRITS itself must now draft rules guiding safe, workable AI rollout across public bodies. The proposal heads next to the House of Representatives.
Office of Management and Budget — Memorandum M-25-21 pushes federal use of AI across US agencies, replacing M-24-10 with fresh demands: a named chief AI officer, an AI governance board, and risk controls on high-impact systems. Innovation now sits alongside stronger federal AI governance and public trust as core aims.
Office of Management and Budget — Memorandum M-25-22 from the Office of Management and Budget replaces Biden-era M-24-18 as United States AI policy for executive agency tenders released after October 1, 2025. Buyers must assess whether a purchase counts as high-impact AI. These AI buying rules demand clauses on performance validation, interoperability and government data rights, and ban training commercial AI with non-public government data.
Superintendencia de Competencia Económica (SCE) — Resolution SCE-DS-2025-12 gave Ecuador's Superintendencia de Competencia Económica (SCE) its first AI ethics code public sector reform, an addition to its existing Code of Ethics that sets out how the regulator governs artificial intelligence use inside a government body. Human review must accompany every automated task, guided by five values -- openness, responsibility, fairness, protection and traceability -- while any system threatening basic rights or producing unfixable harm stays banned outright. By October 2025 the agency had gone further still, issuing a Guide for Auditing AI Use and becoming Latin America's first government institution to build formal algorithmic auditing in government into its oversight structure.
Government of the Czech Republic — Czechia's Czech AI policy now channels CZK 19 billion through an Action Plan tied to the National AI Strategy 2030, run by the Ministry of Industry and Trade. Seven areas share the money: research, education, industry, public administration. A national artificial intelligence roadmap this size marks the country's biggest implementation step yet.
Comunidad Autónoma de Galicia — Galicia's Ley 2/2025 stands as the regional AI law Spain now counts as its first, a single rulebook steering how artificial intelligence gets built and used by government bodies and businesses alike within the community. Under this AI governance framework Galicia now requires civil servants and ordinary residents to receive training in understanding AI systems. The statute folds an older blueprint, the Estrategia gallega de inteligencia artificial 2030, into enforceable legislation.
Ministry of Transport, Communications and Information Technology — Oman's Ministry of Transport, Communications and Information Technology (MTCIT) rolled out the General Policy for the Safe and Ethical Use of Artificial Intelligence Systems, branded the National AI Policy 2025, setting out Oman AI regulation built on humanity and society, fairness and inclusion, and accountability. This national AI governance framework has MTCIT tracking public-sector adherence, reporting to the Council of Ministers, while regulators enforce it sector by sector.
President of the Republic of Belarus — Presidential Decree No 135 ranks 'digital technologies and artificial intelligence' first of five national priority directions for scientific, scientific-technical and innovative activity, setting Belarus AI policy for 2026-2030. Every state scientific programme and the State Innovation Development Programme must steer public R&D funding toward AI priorities and kindred digital technologies.
Romanian Parliament (Senate) — Romania's Senate has registered legislation setting rules for AI regulation in sensitive fields such as healthcare, justice, and public administration, framed around the responsible use of artificial intelligence. Courts would be barred from letting algorithms decide cases, doctors must sign off on any AI-generated diagnosis, and consent becomes mandatory before AI can build political profiles; manipulative deepfakes aimed at swaying public opinion are banned outright. Public bodies deploying such systems face transparency duties and periodic audits, while breaches touching fundamental rights risk fines reaching 100 million lei (20.1 million euros) plus two to five years' imprisonment.
Office of the Privacy Commissioner for Personal Data — The Office of the Privacy Commissioner for Personal Data issued the Checklist on Guidelines for the Use of Generative AI by Employees, Hong Kong AI policy helping organisations draft staff rules under the Personal Data (Privacy) Ordinance. Workplace AI governance topics are permitted scope, data security, lawful and ethical use, personal data privacy and breach penalties.
Office of the Privacy Commissioner for Personal Data (PCPD) — Hong Kong's PCPD released a generative AI checklist for employers: the Checklist on Guidelines for the Use of Generative AI by Employees. It helps firms set internal rules on lawful staff use of these tools under the Personal Data (Privacy) Ordinance, spanning data protection, security and penalties. This AI use policy for employees came with a new AI Security Hotline.
Government of Bermuda — Bermuda's AI policy for public-sector use requires human review for any automated decision touching rights or service access, alongside adherence to the territory's data protection and freedom-of-information rules. Its formal title, the Artificial Intelligence Policy, covers all government employees and consultants using AI tools for official work. A new AI Governance Sub-Committee, part of the government AI governance framework, will monitor rollout and compliance.
Federal Ministry of the Interior and Community (BMI) — Germany's BMI now issues nationwide guardrails covering AI in federal administration, via its 'Leitlinien für den Einsatz Künstlicher Intelligenz in der Bundesverwaltung'. Agencies gain shared principles for rolling out such systems safely and consistently — part of broader traction for AI governance in the public sector.
Ministry of Information, Communications and the Digital Economy — Under the banner Kenya AI policy stands the Kenya National Artificial Intelligence Strategy 2025-2030, steering developers, tech firms and AI-using industries for half a decade. Economic expansion, social inclusion and fresh innovation anchor its goals, echoing a broader national AI strategy movement worldwide. Executive policy status now applies nationwide.
Commonwealth of Virginia — Virginia now treats fabricated or altered artificial-intelligence material as grounds for defamation claims, under a fresh AI content regulation named HB 2124, the Synthetic Digital Content Act. Anyone who fraudulently misuses such content commits a Class 1 misdemeanour, and people portrayed without consent can pursue civil lawsuits against the wrongdoer. Virginia's Attorney General must convene a group to examine how statutes against AI-generated deepfakes get enforced.
Government of Norway — Nkom (Norwegian Communications Authority) leads Norway AI regulation, appointed as coordinating supervisory body over EU AI Act compliance for Norway's Ministry of Digitalisation under Karianne Tung. Firms, consumers and agencies gain guidance from Nkom, while Datatilsynet and Digdir join a shared sandbox advancing AI oversight nationwide.
Swiss Federal Council — Three pillars anchor Switzerland AI policy under the KI-Teilstrategie, the Federal Chancellery's Strategy for the Use of AI Systems in the Federal Administration, briefed to the Swiss Federal Council: sharpening staff skills, winning public confidence through lawful, ethical practice, and lifting output by automating repetitive tasks. A subsequent plan layers on eight concrete measures with a 2026-2028 roadmap, extending AI regulation uniformly across federal agencies.
Swedish Government — A new law would let Sweden's police run AI facial recognition police checks in real time against people in public spaces, part of wider biometric remote identification powers under memorandum Ds 2025:7. Officers would normally need a court's or prosecutor's sign-off first, except in urgent cases, where approval can follow within 24 hours. Swedish consultation on the plan produced a formal government proposition in March 2026.
President of Azerbaijan — A presidential decree signed by Ilham Aliyev put the Azerbaijan AI strategy into effect for the Republic of Azerbaijan, running through 2025-2028. Five focus areas structure the plan — regulatory governance, education and workforce skills, infrastructure and data systems, innovation and research, plus cooperation abroad — while each ministry must appoint a deputy minister handling digitalisation and AI matters. Ministerial accountability thus becomes the backbone of the country's artificial intelligence policy.
President of the Republic of Azerbaijan — Azerbaijan AI policy now runs on a national strategy for 2025-2028, approved by presidential decree from President Ilham Aliyev for the Republic of Azerbaijan. It sets goals across government services, the economy, schooling, data and computing capacity, workforce skills, and safety and ethics oversight, backed by a coordinating body and an action plan to 2028. The approach treats artificial intelligence adoption in Azerbaijan as a cross-sector priority rather than a single-agency effort.
National Commission for the Control of Personal Data Protection (CNDP) — Morocco's CNDP ruled that AI and personal data protection falls under existing Law 09-08. No new legislation is needed: the same safeguards already cover automated processing. This guidance on AI data privacy extends current duties to AI tools.
Governorate of Vatican City State — Vatican City State has stood up a five-member Commission on Artificial Intelligence for AI governance, created under Decree N. DCCII containing the Guidelines on Artificial Intelligence by Sister Raffaella Petrini. Seats went to the Legal Office, the Directorate of Telecommunications and Information Systems, and the Directorate of Security and Civil Protection Services. Ahead: drafting AI regulation, weighing AI deployments, warning of hazards, and issuing twice-yearly findings on AI's reach.
Ministère de la Transition Numérique et de la Digitalisation (Ministry of Digital Transition and Digitalisation) — Côte d'Ivoire's government adopted a national AI policy, SNIA 2030, resting on skills and infrastructure, regional and social inclusion, and ethical governance. It creates a National AI Committee, Safe IA label, National AI Agency (ANIA) and AI Hub, prioritising agriculture, health and education. This AI strategy sets out structures, not funding.
Government of Spain — Spain's government greenlit an AI governance bill aligning domestic policy with the EU AI Act, promoting ethical, inclusive and beneficial use of artificial intelligence. New digital rights let authorities pull AI systems from the Spanish market after a serious incident, backed by fines from €7.5M to €35M for banned practices. This marks an early step toward a broader AI regulation framework.
Conselho Nacional de Justiça (CNJ) — Binding AI rules now cover Brazil's judiciary: Resolução CNJ nº 615 governs AI in the judiciary nationwide, effective 14 March 2025. Judges and staff must guarantee fairness, human oversight, algorithmic auditability and data protection whenever AI tools come into play, watched by the freshly created Comitê Nacional de Inteligência Artificial do Judiciário. Judicial AI governance standards were subsequently revised through Resolução CNJ nº 674 on 25 March 2026.
Saeima (Parliament of Latvia) — A new legal basis for AI regulation in Latvia arrives via the Artificial Intelligence Centre Law, passed by the Saeima, which also stands up a national AI Centre due to open by 31 March 2025. Tasked with linking government, industry and universities, the body will foster AI skills, manage risk and may operate a sandbox for trialling AI systems under real conditions. Its founders — VARAM, the Ministry of Economy and the Ministry of Defence — bring in private companies and academic institutions as partners.
Administrative and Recruitment Organisation of Iran — Every executive body bound by Iran's Civil Service Management Law must now deliver AI training for government employees, under Circular No 116958 (14/12/1403) from the Administrative and Recruitment Organisation. Coverage includes basic concepts, practical tools, governance issues and long-term strategy, building AI literacy among staff and showing where it fits into daily paperwork. Compliance updates are compulsory, with room for extra specialised sessions if gaps remain.
Treasury Board of Canada — Canada's Treasury Board has unveiled its AI strategy for federal public service work, a 2025-2027 plan from Ginette Petitpas Taylor at the University of Waterloo. The strategy sets up a centre of AI expertise, mandates security and responsible-use rules, builds talent pipelines, and pledges transparency on government AI adoption plans.
Government of Ireland — Eight bodies will share EU AI Act implementation in Ireland instead of one super-regulator: the Central Bank of Ireland, the Data Protection Commission and the Health Products Regulatory Authority among them. Each already polices AI regulation in its own sector. A new AI Office of Ireland coordinates centrally from August 2026.
National Police Chiefs' Council — AI policing gets a national roadmap in England and Wales: the National Police Chiefs' Council's AI Strategy 2025 seeks global leadership in responsible artificial intelligence for stopping, spotting and probing crime. Governance, ethics and milestones to 2027, including a national AI lab, support the goal.
Cabinet of Ministers of the Government of National Unity — Libya's Cabinet of Ministers created the National Programme for Artificial Intelligence and Digital Transformation under Decision No. 40 of 2025, dated 28 February, tasking it with AI in public administration. The new body stands as its own legal entity with a separate budget, sitting inside the Government Communication Centre and reporting to the Minister of State for Communication and Political Affairs. Its brief: speed up public-sector workflows, grow state digital skills, shape AI rules, and push big-data analytics into how government decisions get made — reflecting broader government AI policy goals.
Ministry of Digital / National Digital Department (JDN) — Malaysia now has a 158-page rulebook for AI adoption in the public sector, from the Ministry of Digital. Gobind Singh Deo said the Public Sector AI Adaptation Guidelines cover ethics, risk controls, roles and self-check templates for agencies. Transparency and accountability anchor government agencies adopting AI responsibly.
National Academy of Sciences of Belarus — Belarus's National Academy of Sciences (NAS) has opened the Scientific Artificial Intelligence Center (AI-center), an artificial intelligence research centre built from the United Institute of Informatics Problems' digitalisation office, with a partner firm from the China-Belarus Industrial Park 'Great Stone'. Now the country's main AI research and development hub, it covers cybersecurity, supercomputing, neural network coding, image recognition and language-processing systems, serving as Belarus's core base for both theoretical and applied AI work.
Asamblea Legislativa de El Salvador — El Salvador AI regulation now flows from one statute, the Law for the Promotion of Artificial Intelligence and Technologies. The National AI Agency (ANIA) leads enforcement, backed by the National Registry for AI Development, Innovation and Application. Registered firms gain legal shields under this AI policy framework.
Ministru kabinets (Cabinet of Ministers) — Latvia AI regulation now hands lead authority to the Ministry of Smart Administration and Regional Development (VARAM) for the EU AI Act. Three bodies share market oversight — Consumer Rights Protection Centre, State Data Inspectorate and Bank of Latvia — while the Ombudsman covers fundamental rights. Cabinet approved this EU AI Act implementation plan.
Presidency of the Bolivarian Republic of Venezuela — A three-pillar national AI plan Venezuela unveiled under President Nicolás Maduro builds specialised talent, sovereign computing infrastructure for domestic models, and capacity for the ALBA AI agency. Branded the National Artificial Intelligence Plan for Sustainable Development, it launched 25 February 2025 for transformation via innovation, now covering education, health, hydrocarbons, and banking and finance under this sovereign AI development strategy.
Government of Hungary — Resolution 1028/2025 gives Dr László Palkovics Hungary's AI policy brief as mesterséges intelligenciáért felelős kormánybiztos (government commissioner for artificial intelligence). Remit: defence, health, trade, education, energy, transport, public administration, science, law enforcement, agriculture, financial sector, economy, industry and national security. AI governance duties include founding and chairing the Hungarian AI Council.
APDP (Autorité de Protection des Données Personnelles) — Monaco's Personal Data Protection Authority joins 61 other regulators in a joint statement on AI image generation tools producing lifelike, identifiable images and videos without consent. Signatories demand stronger safeguards, real transparency, working takedown routes, and child protections against this AI-generated deepfake harm.
Congress of the Dominican Republic — Congressman Tobías Crespo filed a Dominican Republic AI regulation bill setting up CONIA, the National Council for Artificial Intelligence, to enforce codes, appeals and sanctions. Bans cover AI governance bill terms like military and security uses, social scoring, biometric categorisation and manipulation.
Government of Namibia — Namibia AI policy now runs through a new National AI Council driving the National Artificial Intelligence Strategy, backed by Technical Working Groups for healthcare, education, agriculture and mining. The goal: regional leadership in AI adoption and innovation.
Cámara de Diputados de la República Dominicana — Fuerza del Pueblo lawmaker Tobías Crespo has lodged Dominican Republic AI regulation legislation with Congress, proposing a National Council for Artificial Intelligence, CONIA, to police how AI is built and deployed nationwide. His bill, part of a wider push for AI governance legislation, would bar the armed forces from using AI and stop agencies from scanning crowds with live facial recognition, reading emotions algorithmically, or ranking citizens through social scoring.
Departamento Nacional de Planeación (DNP) / CONPES — CONPES 4144 sets Colombia AI policy: six pillars spanning ethics, infrastructure, research, talent, risk and adoption, running to 2030. The National Council for Economic and Social Policy approved the document, giving the country a long roadmap for regulating and rolling out artificial intelligence across public and private life.
Swiss Federal Council — Ratifying the Council of Europe's AI Convention has won out over a unified AI regulation in Switzerland, the Federal Council confirmed. Field-specific rules will apply to healthcare and transport, while wider protections stay confined to fundamental-rights topics like data protection. Departments must deliver a consultation draft by the end of 2026, so Swiss artificial intelligence law seems improbable before 2028.
Ministry of Digital Affairs — Poland's Ministry of Digital Affairs has reworked its draft Act on Artificial Intelligence Systems after weighing over two thousand consultation comments. This Polish AI regulation creates the Commission for AI Development and Safety (KRiBSI) as watchdog, lets micro, small and medium enterprises join sandboxes free, and adds individual opinions under the AI Systems Act, giving firms certainty pre-deployment.
Subsecretaría de Gobierno Digital, Provincia de Buenos Aires — Argentina's Buenos Aires province now enforces generative AI guidelines public sector-wide under Resolution 4/2025. Staff must tag anything an AI system generates, ensure a person reviews it before use, and never enter sensitive information into these tools; machines cannot rule on matters touching individual rights. The Dirección Provincial de Innovación Digital will oversee adherence under these government AI guidelines and provide training for personnel.
National Data Protection Commission (CNPD) — AI regulation Luxembourg materials from CNPD brand most unacceptable-risk systems under the EU Artificial Intelligence Act unlawful, arguing these prohibited AI systems generally violate GDPR too. Flagged practices span social scoring, predictive policing, emotion tracking on the job and harvesting facial images without consent, leaving Luxembourg-based deployers exposed to inquiry.
Gouvernement français (Comité interministériel de l'intelligence artificielle) — Under France AI policy, Prime Minister François Bayrou has unveiled a third stage of national AI strategy titled 'Faire de la France une puissance de l'IA'. Agreed at an interministerial gathering, the blueprint sets four aims: expanding computing and data-centre capacity, drawing skilled workers via dedicated visas and research chairs, embedding generative AI tools inside every government ministry, and spreading adoption across the wider French economy through the Osez l'IA scheme. This follows two earlier rounds of the same push to make the nation a leading force in artificial intelligence.
French Government (Ministry of Public Action and Ministry for AI and Digital) — Laurent Marcangeli and Clara Chappaz revealed France's AI in public sector plan on 6 February 2025, reaching Mayotte as a French department. Backed by an interministerial committee under the Prime Minister, the pair committed to faster adoption spanning every ministry, benefiting staff and citizens through new training and tooling. Building on the 2018 national strategy, the announcement signals renewed AI adoption across government.
Personal Data Protection Office (UODO) — The Polish Personal Data Protection Office (UODO) advises extreme caution before anyone uses DeepSeek R1, a Chinese AI chatbot. UODO's AI chatbot data protection concern centres on storage: information may sit in China, within reach of Chinese authorities. The regulator is assessing, on chatbot privacy grounds, whether DeepSeek's data processing meets EU requirements.
Ministry of Communication Technologies — Tunisia's Ministry of Communication Technologies will open its national AI strategy to public consultation by end-February 2025, Sofiene Hemissi said. Thirty diaspora specialists joined local experts in drafting the Tunisia AI policy, spanning infrastructure, skills, legal frameworks, governance and ecosystem-building. The wider artificial intelligence roadmap, shaped by these 30 contributors, is still being finalised.
Federal Executive Council — Nigeria now has an AI oversight body Nigeria calls the National Artificial Intelligence Trust, set up by President Bola Tinubu's Federal Executive Council. Ten (10) AI specialists and two ministers sit on this panel, charged with pooling funds and driving national AI strategy across Nigeria's leading economic sectors.
Government Chief Digital Officer (GCDO) — Under the Responsible AI Guidance for the Public Service: GenAI, New Zealand's Government Chief Digital Officer requires agencies deploying generative AI in government to set up governance boards spanning privacy, security, legal and Māori representation. Human oversight for high-stakes decisions, proportional risk assessments and Māori data sovereignty are compulsory. This replaces interim July 2023 guidance on AI in the public sector.
Belgian Federal Government — The Belgian Institute for Postal Services and Telecommunications now holds the AI Act market surveillance authority role, fixed by the Government Declaration and reaffirmed in the 2025-2029 Federal Government Agreement. This AI Act enforcement authority spreads fundamental-rights supervision across twenty-one bodies under Article 77, the highest count among EU member states.
Department of Internal Affairs (Government Chief Digital Officer) — New Zealand now has a formal method for rolling out AI safely across government, set by Paul James under the Public Service AI Framework. It covers every kind of AI in public services, part of a wider effort tied to the National AI Strategy. Public sector AI governance gains a fixed reference point.
Government of Cyprus — Cyprus AI Act regulation now runs through three bodies the Council of Ministers named to meet EU requirements. The Deputy Ministry of Research, Innovation and Digital Policy is national coordinator on the European AI Board. The Commissioner of Electronic Communications holds the Notifying Authority, Market Surveillance Authority and Single Point of Contact roles; the Commissioner for Personal Data Protection is Market Surveillance Authority for high-risk AI systems, completing this AI Act implementation authorities structure.
Council of Ministers — A new panel now drives Cyprus AI policy: the National AI Taskforce, cleared by Cyprus's Council of Ministers on 22 January 2025 under the Chief Scientist for Research, Innovation and Technology. Members chosen for AI expertise hold five-year posts, guiding national AI strategy, public adoption and innovation.
Cabinet of Ministers of the Kyrgyz Republic — Resolution No. 29 gives Kyrgyzstan a National Council on AI Development, an advisory body under the Cabinet of Ministers steering national AI strategy and linking ministries on implementation. Officials want this artificial intelligence development council to open the door for AI tools across priority economic sectors, keep projects running smoothly, and grow scientific and educational skills in the field. It came into force on 31 January 2025.
Ministry of Communications and Information Technology, National Council for Artificial Intelligence — The National Council for Artificial Intelligence, under the Minister of Communications and Information Technology, issued the National Artificial Intelligence Strategy 2025-2030 second edition, unveiled by President Abdel Fattah El-Sisi. Egypt AI policy rests on six pillars for national AI planning: skills development, ecosystem, infrastructure, data infrastructure, technology development and governance.
Office of the Australian Information Commissioner — Feeding personal information into publicly accessible generative AI tools now draws a caution from the OAIC, whose updated AI privacy guidance covers commercial AI products broadly. Before adopting any such system, organisations are urged to run due diligence checks first. This sits within Australia's expanding AI regulation regime.
Seimas of the Republic of Lithuania — Lithuania AI regulation gained a legal footing when the Seimas passed Act XV-105, aligning the Law on Technology and Innovation with Regulation (EU) 2024/1689. The Innovation Agency now serves as notifying authority for AI conformity assessment, a statutory sandbox for testing AI systems was established, and start-up classification thresholds were narrowed. Most articles applied from 1 April 2025.
Politburo of the Communist Party of Vietnam (Bộ Chính trị) — Vietnam's Politburo formed a 19-member Central Steering Committee for the Development of Science, Technology, Innovation and Digital Transformation, chaired by Party General Secretary To Lam, reflecting a shift in Vietnam AI policy. Debuting 13 January 2025, it drives breakthrough science, technology and digital transformation measures, naming AI research and development a priority. It also oversees Resolution 57-NQ/TW, fixing AI as a strategic technology under the country's wider AI development and digital transformation strategy.
Government of the Republic of Serbia — Adopted by Serbia's government, the Strategy for Development of Artificial Intelligence in Serbia 2025-2030 replaces the country's 2020-2025 predecessor and drives Serbian AI policy toward stronger legislation, education funding, innovation, infrastructure and broader public-sector use of AI. Wider AI deployment across government agencies sits alongside efforts to match global technology trends and support the 'Leap into the Future 2027' programme.
EU AI Act bans in force since 2 February 2025, and unaffected by the 2026 delay to high-risk rules, cover predictive policing based only on profiling or personality traits and facial recognition databases built by untargeted internet or CCTV scraping. Police use of live remote biometric identification in public places is also barred outside narrowly defined purposes.
Personal Data Protection Agency of Bosnia and Herzegovina — Bosnia and Herzegovina's Personal Data Protection Agency now flags which processing activities demand an AI data protection impact assessment, targeting automated and algorithmic decision tools. Triggers include extensive profiling, decisions made without human input, marketing or profiling of children's records, bulk handling of sensitive categories, biometric and genetic identifiers, and public-space surveillance. Staff monitoring software and background-check record matching complete the AI risk governance obligations.
Government of Luxembourg — Luxembourg's Bill No. 8476 sets out national authorities for EU AI Act implementation. The CNPD becomes the default market surveillance body, notified body for high-risk systems in law enforcement and immigration, and sandbox operator; ALIA takes on generative AI transparency oversight, forming the basis of AI Act compliance in Luxembourg. Submitted December 2024, the draft spent 2025 before the Media and Communications committee, drawing input from 12 bodies including the CNPD and Council of State.
Government of Vanuatu — Vanuatu's Digital Transformation Act No. 42 of 2025, in force from 17 December 2025, applies Vanuatu AI regulation through a permit regime for ICT operators, carving out a distinct category for artificial intelligence and AI-linked data services. Oversight sits with the Department of Communications and Digital Transformation, which gains authority to draft national AI frameworks, while a newly formed National Digital Steering Committee vets major cybersecurity, ICT and AI strategies before they go to the Council of Ministers. The law reflects governance built through broader digital statutes rather than a standalone AI-specific act, with annual permits required of all providers, including artificial intelligence in government services.
Ministry of Post and Telecommunications (MPTC) — Cambodia's Ministry of Post and Telecommunications has drafted a National Artificial Intelligence Governance Framework, an early step in shaping Cambodia AI regulation alongside the National AI Strategy. The AI governance framework plan sets expert consultations for July–August 2025, ministry talks from September 2025, and submission to the Digital Government Committee in November 2025.
Data Protection Commissioner (Komisioneri për të Drejtën e Informimit dhe Mbrojtjen e të Dhënave Personale) — Guidance No. 05/2025 on the Processing of Personal Data by Competent Authorities turns Law No. 124/2024 into working practice across Albania's police, prosecution and judicial bodies, zeroing in on AI-driven facial recognition law enforcement activity. The Data Protection Commissioner now insists on impact assessments, logged activity trails and internal whistleblowing channels wherever behavioural profiling or biometric identifiers come into play. Justice-system agencies also face fresh curbs on automated decision-making, widening scrutiny of biometric data processing by police.
Legislative Assembly of Costa Rica — Costa Rica AI regulation cleared a key hurdle as the Legislative Assembly's Human Rights Committee backed Expediente 23.919 with a favourable majority opinion (dictamen afirmativo). Officially named 'Ley para la Promoción Responsable de la Inteligencia Artificial en Costa Rica', this AI governance law envisions ethics-driven oversight for every stage of machine intelligence work nationwide — spanning research, design, construction and eventual rollout — anchored by a proposed watchdog, the Interinstitutional Commission for the Development of Artificial Intelligence (CIDIA), alongside a standalone, autonomous panel of technical and ethical advisers. No plenary ballot had been cast as of September 2026.
Ministry of Transport, Telecommunications and Artificial Intelligence Systems — Equatorial Guinea's Ministry of Transport, Telecommunications and Artificial Intelligence Systems is drafting an Artificial Intelligence Law (Anteproyecto de Ley de Inteligencia Artificial) under Minister Honorato Evita Oma, part of a broader Equatorial Guinea AI regulation push. Six pillars underpin the text: safeguarding national identity, embracing linguistic and cultural diversity, protecting local data and traditional knowledge, pursuing sustainable growth, encouraging training and innovation, and honouring global legal and ethical norms. A proposed autonomous body, ANIAD (National Artificial Intelligence and Data Authority), would provide AI governance oversight, with the effort tied to the African Union Agenda 2063 and the Malabo Convention 2014.
Ministry of Information and Communication Technology — Namibia's Ministry of Information and Communication Technology (MICT) has produced a draft Artificial Intelligence Bill setting out how AI should be used responsibly and safely across government and industry, a step officials describe as central to this Namibia AI regulation drive. Speaker of the National Assembly Saara Kuugongelwa-Amadhila linked the plan to a broader effort at overhauling public administration and curbing hazards from emerging digital tools, and ICT Minister Emma Theofelus later told Parliament in September 2025 that drafters are matching the Bill against UNESCO's guidance on AI ethics as this AI governance legislation takes shape.
Kişisel Verileri Koruma Kurumu (KVKK — Personal Data Protection Authority) — Türkiye's KVKK (Personal Data Protection Authority) has issued AI data protection guidance for anyone building, supplying or deciding with such systems. Publication No. 76, 'Yapay Zekâ Alanında Kişisel Verilerin Korunmasına Dair Tavsiyeler' (Recommendations on Protection of Personal Data in the Field of Artificial Intelligence), calls for lawfulness, transparency, minimal data use and bias safeguards from the outset, plus human oversight and accountability — the essence of personal data protection in artificial intelligence. High-risk projects need a Privacy Impact Assessment, and anyone hit by an automated decision keeps the right to challenge it.
Ministry of Communication and Information Technology — Tanzania will issue a national guideline for responsible AI use, part of a wider Tanzania AI regulation push confirmed by Mohammed Khamis Abdulla. The rules are meant to shape how organisations deploy the fast-moving technology nationwide. A planned Nala, Dodoma institute for digital and AI technologies would anchor training and support, giving this artificial intelligence policy effort a physical base.
Ministry of Science and Technology — Myanmar's Ministry of Science and Technology met in Nay Pyi Taw on 18 February for a fourth session on a National AI Strategy and National AI Policy spanning 2025–2030. Myo Thein Kyaw said the aim is weaving artificial intelligence into growth sectors while lifting public grasp of AI ethics and standards — shaping Myanmar AI policy grounded in the ASEAN Guide on AI Governance and Ethics. This national AI strategy for Myanmar awaits approval.
Ministry delegated for Artificial Intelligence and Digital, Government of France — The French government's Osez l'IA plan sets 2030 targets for AI adoption in businesses: 50% of very small businesses, 80% of SMEs and 100% of large groups. Awareness comes from 615 AI Ambassadors in 20 regions, alongside training for 15 million professionals by 2030 and company support such as Data AI diagnostics and AI Accelerators. Business AI uptake measures apply France-wide, Martinique included.
Central Agency for Information Technology (CAIT) — Kuwait AI policy takes shape as CAIT unveils a draft national AI strategy spanning 2025 to 2028. Five sectors anchor the rollout: government, healthcare, energy, education and transport, with Kuwait pitched as a regional AI leader by 2028. The blueprint ties into the country's wider Kuwait Vision 2035 agenda.
Agência de Protecção de Dados (APD) — A new artificial intelligence chapter sits inside Angola's revised data protection law, Lei n.º 22/11, following public consultation by the Agência de Protecção de Dados (APD) that ended in April 2025. Under this Angola AI regulation, Article 36 outlaws algorithmic social scoring, Article 37 lets anyone harmed by an AI system demand explanation, dispute the outcome and seek human review, and Article 38 makes producers, suppliers and operators liable, all part of the country's artificial intelligence data protection law reforms. Extraterritorial reach also widens, catching foreign processors handling data tied to people inside the country.
Council of Ministers of the Republic of Belarus — Belarus's Council of Ministers approved the Concept of Ensuring Sovereignty of the Republic of Belarus in the Field of Digital Development until 2030 under Resolution No 1074, the country's first coordinated strategy on digital sovereignty. It sets principles for national ICT control, faster rollout of new technology, and modernisation of government, economic and social services, folding AI in public administration into this wider push. As a foundational document it lays out a digital independence policy framework rather than detailed implementation steps.
Government of Luxembourg (tabled by Minister Elisabeth Margue in the Chamber of Deputies) — Draft Law No. 8476 enacts the EU AI Act (Regulation 2024/1689) at national level, part of AI regulation in Luxembourg, naming the CNPD lead market watchdog and sandbox operator. The CSSF and Commissariat aux Assurances cover financial and insurance oversight, with penalties and redress under EU AI Act implementation also fixed. As of July 2026 the bill still awaited committee approval, after opinions from the Council of State and other bodies.
Parliament of Cameroon / President of the Republic — Cameroon AI regulation begins with Law No. 2024/017, which President Paul Biya promulgated as the nation's first full statute on personal data. Under Article 44, data subjects may oppose any decision made wholly by automated processing or profiling, whatever its effect, so these automated-decision rules reach further than the GDPR. A new regulator, the Personal Data Protection Authority, supervises compliance; controllers' 18-month adjustment period ended 23 June 2026.
Parliament of Albania (Kuvendi i Republikës së Shqipërisë) — Automated decision-making AI regulation in Albania runs through Law No. 124/2024, replacing the 2008 statute from 1 February 2025. Article 20 echoes GDPR Article 22: a purely machine-driven ruling, profiling included, cannot stand alone if it carries serious legal weight, absent human oversight. Article 53 extends this AI profiling safeguard to policing, blocking automated bias tied to sensitive data. Breach fines reach 2 billion Albanian lek or 4 percent of global turnover.
National Privacy Commission — Advisory No. 2024-04 from the National Privacy Commission applies the Data Privacy Act of 2012 to AI systems processing personal data, requiring transparency, accountability and fairness from controllers. Automated decisions must carry privacy impact assessments plus a human intervention mechanism. The rules also ban AI washing, barring firms from overstating their AI-driven automated decision-making capabilities.
Pontifical Commission for Vatican City State — Vatican City State's Pontifical Commission adopted Decree N. DCCII, setting out Guidelines on Artificial Intelligence effective 1 January 2025. The rules shape AI governance in Vatican City State on data protection, non-discrimination and human oversight, with sector rules spanning health, heritage, courts and security. A five-member Commission on Artificial Intelligence within the Governorate will oversee this AI ethics framework.
Congresso Nacional (Senado Federal) — This Brazil AI regulation, Senator Rodrigo Pacheco's Marco Legal da Inteligência Artificial (Bill PL 2338/2023), advanced from the Senate to the Chamber of Deputies on 17 March 2025, where it awaits a vote. Pacheco first tabled the text on 3 May 2023, setting out duties, rights and oversight rules for building and deploying AI systems nationwide. Provisions cover tiered risk categories, disclosure obligations, algorithmic impact reviews and mandatory human supervision — a broad AI governance legislation package.
Ministarstvo pravosuđa, uprave i digitalne transformacije (MPUDT) — Croatia now has seven bodies handling AI fundamental rights oversight. The Ministry of Justice, Public Administration and Digital Transformation listed them to the European Commission under Article 77(2) of the EU AI Act: the Ombudswoman, the Ombudswoman for Children, the Gender Equality Ombudswoman, the Ombudswoman for Persons with Disabilities, the Personal Data Protection Agency (AZOP), the State Electoral Commission and the Agency for Electronic Media. Each authority is enforcing rights protections against high-risk AI systems.
Ministry of Digital Government (Digitaliseringsministeriet) — Copenhagen commits DKK 62.5 million through 2027 to Denmark's AI strategy: four projects and three values for trustworthy machine intelligence. These cover a Digital Taskforce for AI, a centre on AI's societal impact, a Danish-language model platform, and open-source Danish text data — Denmark's national AI strategy.
High-Level Advisory Committee on Artificial Intelligence under the Prime Minister — Greece national AI strategy now exists in writing: "A Blueprint for Greece's AI Transformation", from the High-Level Advisory Committee on Artificial Intelligence under the Prime Minister, coordinated by the Special Secretariat of Foresight, formed November 2023. Six flagship programmes span government administration, schooling, research, technology and innovation. This is Greece AI policy framework in practice.
Ministry of Technology and Science — Zambia AI policy rests on a five-pillar roadmap from the Ministry of Technology and Science: links and access, dependable data, public trust, entrepreneurial innovation, plus partnerships. Aims include skilled talent, with AI rollout across farming, health, schooling, mining, finance and government. This National Artificial Intelligence Strategy in Zambia treats data and trust as prerequisites for AI adoption.
Agencia de Gobierno Electrónico y Sociedad de la Información y del Conocimiento (Agesic) and Comité Estratégico del Sector Público para la Inteligencia Artificial y Datos — Agesic drafted Uruguay's AI governance strategy, the National AI Strategy for 2024–2030, under Article 74 of Law 20,212, and the Strategic Committee for the Public Sector for AI and Data signed off on it. Instead of leaving artificial intelligence use unchecked, the plan builds rules for ethical, safe and accountable systems alongside stronger technical skills and infrastructure. Officials cast this as a path to inclusive growth, sustainable development and better-run public services.
Asamblea Nacional de Venezuela — Venezuela's National Assembly unanimously cleared the Artificial Intelligence Bill in a first reading, a Venezuela AI regulation framework spanning 57 articles and four chapters on development, applications and rights safeguards. Lawmakers propose a National AI Agency as oversight body under this AI governance framework. Second discussion lies ahead before enactment.
UAE Cabinet — UAE Cabinet approval has given the country its International Stance on AI Policy, guiding UAE participation in global AI governance and setting domestic AI goals within an international cooperation framework. This UAE AI policy rests on six pillars — advancement, cooperation, community, ethics, sustainability, security — jointly developed by the Office of the Assistant Foreign Minister for Advanced Science and Technology and the AI Office, feeding into the nation's stance on international AI regulation.
Department of Communications and Digital Technologies (DCDT) — Twelve strategic pillars now anchor South Africa AI policy: talent development, digital infrastructure, ethical AI and public sector implementation among them, drafted by the Department of Communications and Digital Technologies (DCDT) within its National Artificial Intelligence Policy Framework. This marks only an opening move toward a fuller national policy, intended down the line to underpin binding AI regulations and perhaps an AI Act. Nothing here binds anyone yet — the framework merely maps a route toward future national AI governance.
Ministry of Science, Innovation, Technology and Telecommunications (MICITT) — The Ministry of Science, Innovation, Technology and Telecommunications (MICITT) launched the National Artificial Intelligence Strategy (ENIA), Central America's first AI public policy. Costa Rica AI policy sets seven strategic axes, including ethical AI, research and innovation, talent and skills, smart government, and infrastructure and data, guided by principles of transparency, human dignity, equity, cybersecurity and risk management. The national AI roadmap prioritises 5G rollout and a National Centre of Excellence in AI.
Ministry of Homeland Security and Technology — Maldives cabinet has adopted an AI governance framework Maldives officials call the National AI Masterplan 2025-2035, tasking the National Centre for Information Technology with steering a decade of artificial intelligence policy. The plan, under the Ministry of Homeland Security and Technology, pushes for a standalone oversight body, a multistakeholder advisory council, public AI education drives, and digitised Dhivehi-language datasets. Muizzu's government frames this as capturing AI's economic upside while containing its risks, part of a broader national AI strategy.
President of the Republic of Uzbekistan — Uzbekistan's Presidential Resolution PQ-358, signed 14 October 2024, sets a Uzbekistan AI strategy targeting $1.5 billion in AI-based software products and services by 2030. It also targets a 10% share for AI-based services on the e-government portal, 10 new AI research laboratories, and a top-50 ranking on the Government AI Readiness Index. This Uzbekistan AI policy created the Center for Development of Artificial Intelligence and the Digital Economy under the Ministry of Digital Technologies, with an action plan for 2024-2026.
Ministry of Economic Development and Digitalisation — Moldova's Government staked its stance on AI regulation in a White Paper on Artificial Intelligence and Data Governance, drafted by the Ministry of Economic Development and Digitalisation to steer the nation's tech landscape. Non-binding, it signals intent, paving way for future measures unwritten. Its goal: ethical, inclusive, sustainable AI governance matching EU norms.
Federal Chancellery (Bundeskanzleramt) — Eight rules now shape AI use federal civil servants must follow in Austria's federal public service, set by the Federal Chancellery's KI-Guidelines. Staff keep oversight and final accountability, treating tools as merely supportive, restricted to data protection-compliant inputs, with generated content flagged openly. Outputs face screening for bias and ethics, and sustainability extends AI tools in public service beyond accuracy.
Secretaria de Governo Digital (SGD), Ministério da Gestão e da Inovação (MGI) — Brazil's Secretaria de Governo Digital runs the Núcleo de Inteligência Artificial (NIA) to spread AI adoption in government agencies nationwide. Established by Portaria SGD/MGI nº 6.618 and revised via Portaria SGD/MGI nº 473 of 20 January 2026, it gives ministries shared software, coaching, ethics rules and hands-on backing. This centralises AI governance in government instead of each body building capacity separately.
Ministry of Transport, Communications and Information Technology — The Ministry of Transport, Communications and Information Technology's National Program for Artificial Intelligence and Advanced Digital Technologies upgrades the 2022 Executive Program and anchors Oman AI policy for 2024–2026. Its three pillars cover AI adoption in economic sectors, AI localisation and human-centred AI governance. The national AI strategy spans a National Open Data Platform, National AI Research Centre, AI Studio and Omani language model.
Ministry of Science, Technology and Innovation (MOSTI) — Malaysia's Ministry of Science, Technology and Innovation (MOSTI) issued AIGE, the National Guidelines on Artificial Intelligence Governance and Ethics, a non-binding Malaysia AI governance blueprint for safe, fair and accountable AI use nationwide. Seven pillars underpin it — human benefit, accountability, transparency, inclusivity, privacy and data security, reliability, safety and control, plus fairness — named in the guidelines themselves. It sits inside the National AI Roadmap 2021-2025 and is built for revision over time, tracking wider shifts in AI ethics and governance.
Ministry of Digital Transformation, Innovation and Modernization of Administration — Twelve objectives, thirty measures, five priorities: Mauritania's National Artificial Intelligence Strategy 2024–2029 covers skills, research, data governance, AI ethics, and use across healthcare, education, farming, transport, energy and defence — a national AI policy reaching most industries. The Ministry of Digital Transformation, Innovation and Modernization of Administration leads it; a sub-committee under the Supreme Council for Digitalization oversees it. The OECD AI Policy Navigator lists it active.
Belgian Data Protection Authority (DPA/GBA) — Belgium's Belgian Data Protection Authority released a 23-page brochure on 19 September 2024 under Belgium AI regulation, mapping how the EU AI Act meshes with the GDPR to guide AI compliance frameworks. Titled "Artificial Intelligence Systems and the GDPR - A Data Protection Perspective", it shows where core privacy duties like openness, fair processing, lawful basis, minimal data use, limited purpose and accountability now stretch to cover AI systems, plus fresh or heavier obligations placed on providers and deployers handling personal information.
Saudi Data and Artificial Intelligence Authority (SDAIA) — SDAIA (Saudi Data & AI Authority) launched Saudi Arabia's AI Adoption Framework at the third Global AI Summit: a roadmap for embedding artificial intelligence government-wide, covering maturity assessment, AI offices, and best practices. This national AI adoption strategy also confirmed 23 government entities now have active AI offices.
Government of Andorra — A newly created body, the Data Intelligence Agency (AID), now leads Andorra AI governance under the Government of Andorra's Decree 327/2024. Its remit spans executing the national data and AI strategy, building a shared methodology, and drafting ethical and regulatory rules for artificial intelligence. AID also champions real-world projects blending data analytics with AI, shaping the country's Andorran AI governance strategy going forward.
Cabinet of Sri Lanka (Ministry of Technology) — The Cabinet approved the National AI Strategy (AI Sri Lanka 2028), a five-year Sri Lanka AI policy aimed at a regional AI hub and economic resilience by 2030, with Rs 1 billion of the 2024 budget for initial work. Target sectors are public safety, public sector services, agriculture, finance, education and healthcare. Seven principles, among them inclusivity, transparency and human-centricity, guide this national AI plan, drafted by the Committee on Formulating a Strategy for AI (CFSAI) with UNDP support.
Presidencia de la República de Guinea Ecuatorial — Equatorial Guinea's President established a Secretary of State for Artificial Intelligence Systems post under Decreto Núm. 86/2024, naming José Eko Owono Nve its inaugural holder within the Ministry of Transport, Telecommunications and Artificial Intelligence Systems, a government AI leadership role with formal responsibility for steering national artificial intelligence policy. Decreto Núm. 61/2026 of 23 June 2026 confirmed the office's continuation under successor Dr. Óscar Ondo Ngomo.
Superintendencia de Industria y Comercio (SIC) — Binding guidelines on personal data processing in AI systems now apply nationwide after Colombia's Superintendence of Industry and Commerce (SIC), acting as data protection authority, issued Circular 2. Firms building or operating systems involving AI data processing must justify each use of personal information against tests of suitability, necessity, reasonableness and proportionality, while embedding privacy safeguards from the design stage and proving accountability through documented measures. The directive shifts responsibility onto controllers rather than regulators to police compliance case by case.
Council for Transparency (Consejo para la Transparencia) — Chile's Council for Transparency (CPLT) issued voluntary guidance on algorithmic transparency in automated decision-making, targeting agencies that run automatic or semi-automatic decision tools, through Resolución Exenta N° 372 on 30 August. A follow-up covering automated decision systems transparency arrived 14 July 2025 to guide implementation.
Government of the Republic of Serbia — A decision by the Government of the Republic of Serbia, published in the Official Gazzette on 31 July 2024, gave the country its Council for Artificial Intelligence, a step shaping Serbia AI policy. Prof. Dr. Igor Peško leads the group, whose ranks pull together voices from public administration, universities and business to steer and check progress on the nation's AI development strategy. Beyond advising, members are charged with shaping and reviewing the strategy's action plan and pushing forward legislation and rules touching artificial intelligence.
Council of Ministers (Këshilli i Ministrave) — Albania's Council of Ministers made binding rules for AI use in public sector systems across the Republic of Albania through Decision No. 479, setting methodology and technical standards for artificial intelligence government systems. The central obligation requires human oversight built into each system, so people can review and correct outputs that turn out wrong or cause harm. Unlike voluntary guidance seen elsewhere, this measure carries legal force within Albania's public administration.
Government of the Czech Republic (via Ministry of Industry and Trade) — Czechia adopted its Czech Republic AI policy blueprint on 24 July 2024, naming targets to reach by 2030 across seven linked fields: research, teaching, skills-building, legal and ethical questions, security, commerce, and government operations. Officials behind the National Artificial Intelligence Strategy of the Czech Republic 2030 (NAIS) frame the document as a push to turn the country from a mere buyer of advanced AI systems into a builder of them, a shift also described as pursuing a national AI strategy for Czechia.
Government of the Republic of Kazakhstan — The Government of the Republic of Kazakhstan has launched a Kazakhstan AI policy under Resolution No. 592, setting a five-year national AI strategy through 2029 across data, infrastructure, talent, R&D and legal rules. Targets include training five million people in AI skills, building five domestic unicorns, raising IT exports to $5 billion, and deploying at least 25 AI products yearly across economic sectors. The plan is formally titled the Concept for the Development of Artificial Intelligence.
Agenzia per l'Italia Digitale (AgID) and Presidenza del Consiglio dei Ministri — AgID's Italy AI policy blueprint, the Italian Strategy for Artificial Intelligence 2024-2026, rests on four macro-areas: Research, Public Administration, Business and Training. Gianluigi Greco chaired the panel shaping it; the document adds a monitoring system plus regulatory analysis. It stands as Italy's leading Italian AI regulation reference through 2026.
Data Protection Commission — Ireland's Data Protection Commission issued 'AI, Large Language Models and Data Protection' on 18 July 2024, mapping GDPR and Irish law onto large language model training and use. The guidance addresses AI data protection concerns for organisations building generative AI tools, covering fairness, lawfulness, transparency, storage limitation, data subject rights and automated decision-making.
Agencia de Gobierno Electrónico y Sociedad de la Información y del Conocimiento (Agesic) and Unidad de Acceso a la Información Pública (UAIP) — Uruguay's Agesic agency and the Access to Public Information Unit (UAIP) published the Recommendations on Algorithmic Transparency, algorithmic transparency public sector guidance, on 17 July, built under the Fifth National Open Government Action Plan 2021-2024. The document hands state bodies practical tools for explaining automated decisions, seeking to grow citizen confidence while meeting current access-to-information and disclosure law covering algorithms that public agencies run. This effort pushes AI transparency in government further by tying ethical, accountable system use to legal compliance obligations already in force.
European Parliament and Council of the European Union — Spain AI regulation now runs on Regulation (EU) 2024/1689, adopted 13 June 2024 by the European Parliament and the Council and printed in the EU Official Journal, applying automatically with no domestic transposition step. Outright bans hit unacceptable AI practices, high-risk systems face fresh duties, and each member state must appoint a watchdog — AESIA got that job in Spain. A single bloc-wide text thus decides deployment rules under this risk-tiered EU AI Act.
Government of Romania — Romania AI policy for 2024-2027 rests on six priorities under the National Strategy in the Field of Artificial Intelligence (SN-IA): skills, infrastructure, research, technology transfer, wider societal uptake, and governance. Government Decision HG 832/2024, adopted 11 July, set up an Inter-ministerial Commission for the Coordination of the Implementation, drawing on 34 institutions. It doubles as national AI regulation, reaching government and society alike.
Supreme Council of the Cultural Revolution of Iran — Under Iran AI regulation, the Statute of the National Artificial Intelligence Organisation (NAIO) took effect on 10 July 2024, following a decree from the Supreme Council of the Cultural Revolution and the naming of its first Executive Director. Answering directly to the President of Iran rather than any ministry, this new national AI governance body pulls licensing, compute planning, technical standard-setting and sandbox oversight under one roof, while also driving forward a homegrown Persian large language model. Its founding statute further demands that data stay within national borders, that public-facing tools work in Persian, and that riskier applications like facial recognition undergo ethics review.
Parliament of Egypt (House of Representatives) — A 24-article draft on AI governance in Egypt, put forward by lawmaker Amira Saber, would become the country's first such statute before parliament. Artificial intelligence systems get sorted into risk tiers, eight practices such as deepfakes and biased algorithms face a ban, and public agencies must roll out AI within three years, each ministry and governorate appointing its own AI officer under this artificial intelligence oversight legislation. Cover against cyberattacks and data breaches becomes compulsory too; the bill has yet to clear parliament.
Council of Ministers, Federal Democratic Republic of Ethiopia — Ethiopia's Council of Ministers has cleared the country's first National Artificial Intelligence Policy, passed together with a batch of other bills and draft policies in the same sitting. This Ethiopia AI policy sets priorities spanning health, finance, agriculture and security, anchored in data governance, ethics and skills-building. Oversight of rollout falls to the Ethiopian Artificial Intelligence Institute, named as the lead authority steering AI regulation nationwide.
Asamblea Nacional — Ecuador's National Assembly is reviewing the Proyecto de Ley Orgánica de Regulación y Promoción de la Inteligencia Artificial en Ecuador, filed by legislator Patricia Núñez under process 450889, which would classify systems by four risk tiers modelled on the EU, ban social scoring and subconscious manipulation, and create an AI Ethics Committee for AI regulation in Ecuador. A risk-tiered oversight structure like this remains stuck in committee as of February 2026, with no first-debate report issued.
Supreme Council of the Cultural Revolution of Iran — Iran's Supreme Council of the Cultural Revolution ratified the National Artificial Intelligence Document of the Islamic Republic of Iran on 29 Khordad 1403, launching an Iran AI policy of 14 policies and 11 major goals. A National AI Organisation and Steering Council for AI must now stand up, chasing a top-ten world AI ranking within ten years. Ministries and agencies inherit duties under this national artificial intelligence strategy, covering infrastructure, data, ethics and workforce skills.
Malta Digital Innovation Authority — Malta launched the world's first national AI certification programme in October 2019 through the Malta Digital Innovation Authority's voluntary AI Innovative Technology Arrangement (AI ITA) scheme. Approval follows a risk-based assessment against published objectives on governance, security, auditability and data protection. The ITAS Certification Regulations (Subsidiary Legislation 591.01) fixed fees and legal mechanics for this AI compliance certification route, and the Technology Assurance Sandbox (MDIA-TAS) provides parallel testing for iterative development.
Parlement européen et Conseil de l'Union européenne — This AI regulation, Regulation 2024/1689, better known as the EU AI Act, now gives Brussels a single rulebook for building, selling and running artificial intelligence systems across the bloc. Machines get sorted into bands stretching from negligible danger up to outright prohibition, while firms that supply or operate them answer to watchdogs appointed in each nation for market oversight. Because it binds every member state automatically, French Guiana shares in that AI governance regime simply by belonging to the outermost tier of the European Union.
European Parliament and Council of the European Union — AI regulation Réunion rests on Regulation (EU) 2024/1689, the EU AI Act, passed 13 June 2024 by the European Parliament and Council and printed in the Official Journal on 12 July 2024. Bloc-wide and binding, it treats Réunion as an outermost EU territory like any other member state, grading AI tools by hazard level and fixing duties on providers and deployers: bans for unacceptable-risk uses, disclosure rules, conformity checks. That single AI governance framework now governs deployment across the union.
Ministry of Digital Development, Innovation and Aerospace Industry — Kazakhstan now has a National Ethical Code in the Field of Artificial Intelligence, issued by the Ministry of Digital Development, Innovation and Aerospace Industry, setting out AI ethics in Kazakhstan across six areas: putting people first; safety and reliability; being transparent and explainable; accountability; avoiding unfair bias; and protecting the environment. Signing is optional for state bodies, businesses and researchers, though it may improve odds in government contract bids. This Kazakhstan AI regulation leans on incentives rather than legal duty to encourage buy-in.
Office of the Privacy Commissioner for Personal Data — Hong Kong's Office of the Privacy Commissioner for Personal Data published the Artificial Intelligence: Model Personal Data Protection Framework on 11 June 2024, giving organisations recommendations for AI data protection compliance under the Personal Data (Privacy) Ordinance. It sets out four areas: AI strategy and governance, risk assessment with human oversight, system customisation and management, and stakeholder communication. The guidance targets procurement, implementation and use of AI systems while safeguarding personal data protection in AI systems.
UAE Cabinet / Artificial Intelligence, Digital Economy and Remote Work Applications Office — 12 ethics rules steer how UAE federal bodies design and deploy artificial intelligence, under a Cabinet-approved Charter for the Development and Use of Artificial Intelligence. Safety, bias mitigation, privacy, transparency, oversight, accountability and legal compliance anchor this AI governance UAE framework, kept advisory rather than compulsory. It underpins the UAE AI Strategy 2031, giving agencies a common blueprint for responsible AI deployment.
Council of Ministers of the Republic of Cuba and Ministry of Communications (MINCOM) — Six pillars frame the Estrategia para el Desarrollo y Uso de la Inteligencia Artificial, folded by Cuba's Council of Ministers into the wider Política de Transformación Digital under the Ministry of Communications: ethics and law, workforce skills, applications, government use, research, and public messaging. Binding rules demand safe, explainable, fair, privacy-conscious systems under this AI development strategy Cuba has adopted. Farming, healthcare, schooling, transport, water, tourism, manufacturing and government bodies top the sectors marked for artificial intelligence deployment.
Federal Public Service Policy and Support (BOSA) / Federal Public Service Economy — A Royal Decree created a Steering Committee on Artificial Intelligence (Sturingscomité inzake artificiële intelligentie) within the Federal Public Service Policy and Support and the Federal Public Service Economy, advising the federal government on Belgium AI policy. The body is the sole federal contact point for AI governance, draws members from several ministries, regions, universities, the National Labour Council, BIPT, the Centre for Cybersecurity Belgium and the language communities, and convenes at least three times yearly.
Parlement de la République du Congo — The Republic of Congo's Parliament created the Centre Africain de Recherche en Intelligence Artificielle (CARIA) under Law 14-2024, a public body pursuing AI research policy via advanced study, training and innovation. Council of Ministers records frame its purpose as letting African scientists produce and circulate frontier findings in artificial intelligence, following a human-centred model widening gains while easing development hurdles.
Garante per la protezione dei dati personali — Italy's Data Protection Authority, the Garante per la protezione dei dati personali, issued guidance on web scraping generative AI training, giving preliminary direction on large-scale harvesting of personal information from online sources for this purpose. The note recommends protective measures for operators of Italian websites and platforms to guard against mass collection of personal data for AI models without consent.
General Authority for Communications and Informatics — Libya has a National Artificial Intelligence Policy from the Communications and Informatics Authority, tied to World Telecommunication and Information Society Day, slogan 'Digital Innovation for Sustainable Development'. This Libya AI policy backs ethical AI across government: ethics rules, transparency standards, infrastructure, local skills. Part of a national AI strategy.
Government of Spain, Ministry for Digital Transformation and Public Service — 1.5 billion euros: Spain has pledged that much behind the Artificial Intelligence Strategy 2024, adopted 14 May to refresh the older 2020 ENIA plan for faster-moving generative AI. One pillar builds capacity through supercomputing, the ALIA language models and talent pipelines; a second drives uptake via GobTechLab, Kit Consulting and Kit Digital; a third anchors ethics and oversight under the Spanish AI Supervisory Agency (AESIA). This is the backbone of Spain's AI strategy, ratified by the Council of Ministers, running through 2024-2025.
Seimas of the Republic of Lithuania — Lithuania's Seimas set eleven principles governing public-sector AI governance under Resolution No. XIV-2620: oversight, accountability, traceability, quality control, equality, non-abuse, transparency, protection of human interests. These AI deployment principles bind government bodies, with no stated enforcement mechanism.
Government of Andorra — Andorra's Government has issued a Code of Ethics on Artificial Intelligence, an official guide steering how ministries, companies and the public approach AI ethics code Andorra requires. Principles cover oversight by humans, openness, safeguarding data and answerability, mirroring benchmarks set by UNESCO, the OECD and the Council of Europe. Certain public-sector uses must follow it; everyone else may choose to apply this responsible AI governance framework.
Austrian Data Protection Authority (Datenschutzbehörde, DSB) — Austria's Datenschutzbehörde (DSB) claimed oversight of high-risk AI deployed in policing, courts and electoral processes, marking its debut FAQ on AI data protection compliance. Firms deploying such systems must meet GDPR's lawfulness, transparency, minimisation and accuracy tests, cross-referenced against the EU AI Act. Handling personal data through automated systems now falls under both frameworks simultaneously.
CNIL (Commission Nationale de l'Informatique et des Libertés) — France's data authority, the CNIL, issued its first AI system development GDPR guidance on 8 April 2024, following public consultation. Topics addressed include the relevant legal regime, purpose definition, actor qualification, legal basis, testing, data protection impact assessments, and data protection by design. Coverage extends across all French territory, including Mayotte.
Cabinet Office — Procurement Policy Note 02/24 pushes AI in government procurement toward greater openness across Whitehall. Suppliers must flag any AI role in drafting bids, authorities must probe related risks, and contracts must bind data used to train AI models. Agencies, public bodies and central departments fall within this transparency in AI procurement drive.
Government of Flanders — Belgium's Flanders region has locked in five more years of funding under the Vlaams Beleidsplan Artificiële Intelligentie (Flanders AI Policy Plan), approved 22 March 2024 and covering 2024-2028. Money splits three ways: 14.185 million euros yearly for core research via the Flanders AI Research Programme, backing for firms adopting AI tools, plus ethics, legal and skills work run by the Knowledge Centre Data & Society and Flemish AI Academy. Total spend across the Belgium AI strategy lands near 34 million euros a year.
Digitaliseringsstyrelsen and Datatilsynet — Free GDPR guidance and AI Act risk classification are now open to Danish firms and public bodies through a sandbox jointly run by the Agency for Digital Government and the Data Protection Authority. This Danish AI regulation effort speeds movement from prototype to deployment, backing safer artificial intelligence adoption nationwide.
Federal Chancellery of Austria and Federal Ministry of Finance — Eleven experts now sit on Austria's AI Advisory Board (KI-Beirat), formed to steer federal AI policy on technical, ethical and societal grounds. Tasked with shaping the country's AI strategy, the panel also scans global trends for risks and openings facing Austria. RTR provides its back-office support and runs the linked AI Service Desk.
National Cyber Security Agency — Version 1.0 of Qatar's AI security guidelines arrived from the National Cyber Security Agency in 2024, giving organisations across every sector a voluntary, bilingual Arabic/English playbook titled Guidelines for Secure Adoption and Usage of Artificial Intelligence. Rather than mandating compliance, the document builds practical advice around three pillars — people, process, technology — spanning the full AI lifecycle: strategic alignment and risk assessment feed into data governance and model security, while oversight by humans, risk in the supply chain, ongoing monitoring and response to incidents complete the picture. This secure AI adoption framework carries no penalties for non-compliance, leaving uptake to individual organisations.
Office of the Ukrainian Parliament Commissioner for Human Rights — Ukraine's Ombudsman office, Ministry of Digital Transformation and the EU4DigitalUA project released joint guidance on AI data protection for public bodies and companies. Officials set ethical rules for artificial intelligence privacy compliance, stressing safeguards against rights abuses, as part of the country's Roadmap for AI Regulation.
Ministry of Economic Affairs and Communications — Estonia AI policy now centres on the AI and Data Action Plan 2024–2026, published by the Ministry of Economic Affairs and Communications to guide artificial intelligence across public agencies and businesses. An AI competence centre anchors the framework, which channels funding toward workforce skills, language technology and private-sector innovation while steering how government bodies deploy machine learning tools. This national AI strategy came into force on 8 February 2024 and remains active today.
Department for Science, Innovation and Technology — "Implementing the UK's AI Regulatory Principles: Guidance for Regulators" tells regulators how to apply five cross-sectoral principles within their own remits. UK AI regulation rests on safety, security, robustness, transparency, explainability, fairness, accountability, governance, contestability and redress. Each faced a 30 April 2024 deadline for a written plan on artificial intelligence oversight.
Austrian Parliament (Nationalrat) — Austria's Parliament passed a law reshaping AI regulation Austria-wide, amending the KommAustria-Gesetz and Telekommunikationsgesetz 2021. RTR now hosts the KI-Servicestelle (AI Service Desk), a single contact point guiding businesses and citizens through AI rules, while a KI-Beirat panel of eleven specialists advises the federal government on AI's ethical and societal fallout — an AI governance advisory board, not an enforcement body. BGBl. I Nr. 6/2024 carries the text, published 26 February 2024.
Ruler of Abu Dhabi — Law No. 3 of 2024 created Abu Dhabi's Artificial Intelligence and Advanced Technology Council (AIATC), giving it authority over AI regulation Abu Dhabi-wide. Its brief covers research, infrastructure and funding for artificial intelligence and advanced technology, plus project oversight, placing one body in charge of AI policy oversight across the Emirate.
Commission nationale de l'informatique et des libertés — AI regulation France gained a formal shape when CNIL adopted binding guidance on 18 January 2024, under Délibération 2024-011, tying GDPR to AI system design. Firms must classify systems correctly, define purposes, pick a lawful basis, minimise data, assess risks and embed protection by design. This GDPR AI compliance guidance applies equally in French Guiana.
Commission nationale de l'informatique et des libertés (CNIL) — France's CNIL, under Deliberation 2024-011, released its first framework for AI systems compliant with GDPR, addressing how projects establish legal basis, define purpose, qualify actors, manage data collection, and complete data protection impact assessments. As the French data protection authority, CNIL extends this AI development compliance guidance to Réunion.
Commission nationale de l'informatique et des libertés (CNIL) — France's data protection authority CNIL adopted Délibération n° 2024-011, its first General Data Protection Regulation (GDPR) recommendations for building AI systems. Released 8 April 2024 as Martinique AI regulation covering French territory, it addresses purpose, legal responsibilities, legal basis, impact assessments and data protection by design. Its audience: AI professionals needing rules for AI development.
Department of Enterprise, Tourism and Employment — Ireland AI regulation now has an independent advisory voice: the AI Advisory Council held its opening session with Dr Patricia Scanlon chairing. Dara Calleary founded the panel to give ministers impartial specialist input, seeking public confidence in artificial intelligence built around trustworthiness and human needs. Members will also produce their own AI policy advice tracking new developments, dangers and possibilities.
Saudi Data and Artificial Intelligence Authority (SDAIA) — Saudi Arabia's Saudi Data and AI Authority (SDAIA) has issued a generative AI government policy for public sector bodies: the Generative Artificial Intelligence Guidelines for Optimal Adoption in Government Entities. The framework demands compliance with data classification standards, enforces privacy and security safeguards, bans use of restricted data, requires responsible deployment practices, and mandates that humans review every output before release. Each entity must also create a dedicated data management office or AI governance unit. The move sits within Saudi Arabia's broader drive toward AI regulation, though binding enforcement mechanisms remain unspecified.
Federal Chancellery (Bundeskanzleramt) — Version 2.0 of the Praxisleitfaden Digitale Verwaltung: KI, Ethik und Recht gives Austria's Federal Chancellery a working reference for AI ethics in public administration, covering how officials should plan, design, procure and deploy AI systems lawfully. New material folds in generative AI, quantum AI and EU AI Act duties, paired with a Checkliste für ethische KI in der Verwaltung 2.0 for practical use. A first edition preceded it in November 2023, before this update to AI in government administration tools.
Authority for Info-communications Technology Industry (AITI) — Brunei AI governance policy runs through a 25-member panel at AITI, pooling representatives from government, industry and academia. That team wrote the Guide on AI Governance and Ethics for Brunei Darussalam, ran the public consultation and issued two editions. It endures as the nation's lead voice on AI ethics and governance rules.
Hong Kong Judiciary — Hong Kong's Judiciary introduced its first policy on generative AI in courts, covering judges and support staff. It sets broad principles for AI use by judicial officers in judicial and administrative work, framed around guidance from other jurisdictions and the government's Ethical AI Framework.
European Parliament and Council of the European Union — Mayotte falls under Regulation (EU) 2024/1689 by virtue of its status as a French overseas department inside the European Union, meaning the EU AI Act Mayotte rules apply automatically without local transposition. Brussels structured this AI regulation in Mayotte around four tiers of danger — unacceptable, high, limited and minimal — with unacceptable practices outlawed from 2 February 2025 and duties for general-purpose AI models starting 2 August 2025. High-risk system requirements arrive later still, from 2 August 2026, giving developers and deployers a staggered runway rather than a single compliance deadline.
European Parliament and Council of the European Union — Brussels enforces a single artificial intelligence rulebook, the EU AI Act, now binding across the European Union — the world's first such comprehensive law, in force since 1 August 2024. Martinique falls under its scope automatically, as an outermost region of a member state, alongside every other EU territory. Instead of one uniform standard, the text sorts systems into four bands — unacceptable, high, limited and minimal risk — fixing separate duties on providers and deployers depending on where a tool lands. This tiered method to AI governance ties compliance burden to potential harm rather than treating every deployment alike.
Personal Information Protection Commission — South Korea's PIPC has issued a voluntary, CPO-centred 'AI Privacy Risk Management Model' covering AI systems that touch personal data, spanning design, data gathering, build, rollout, use and shutdown. Organisational and technical safeguards plus ongoing oversight apply throughout. This AI Privacy Risk approach backs wider personal information protection in AI systems, though adoption stays optional.
Department of Communications and Digital Technologies — South Africa's Department of Communications and Digital Technologies launched efforts toward comprehensive South Africa AI policy by publishing the country's first national artificial intelligence policy framework. The document sets initial direction rather than final law, serving as groundwork for future policy development.
Ministry of Digital Transformation of Ukraine — Kyiv favours a phased path to AI regulation in Ukraine: self-regulation first, later a dedicated statute matching the EU AI Act. The Ministry of Digital Transformation's White Paper on Artificial Intelligence Regulation in Ukraine outlines a sandbox, conduct codes, labelling and sector advice, short of binding rules. Ukrainian AI law remains a proposal.
Austrian Federal Government (Federal Chancellery and Federal Ministry for Innovation, Mobility and Infrastructure) — The Federal Government's KI-Umsetzungsplan 2024 sets out 47 Austrian AI policy measures across all twelve federal ministries, extending AIM AT 2030 as an interim report. Eight priority areas include trustworthy AI, resilience and security, climate neutrality, AI literacy, digital administration and governance, favouring short- to medium-term action. This national AI strategy ties each measure to European values and the EU AI Act.
National Assembly of Seychelles — A private member's draft, the Pro-Human Technology Bill, would give Seychelles AI regulation covering job loss from automation, safeguarding children's capacity to build natural intelligence at school, and guaranteeing service access without compelling internet use. It also targets digital forgeries and discriminatory outcomes. Srdjana Janosevic put forward this emerging technology legislation, now stalled in a National Assembly working group ahead of full parliamentary debate.
Federal Government of Somalia / Ministry of Communications and Technology — Mogadishu's Somalia National AI Center (SNAIC), a state-linked body, champions research, teaching and cross-sector partnerships to accelerate Somalia AI policy goals. Somali National University marked its official launch in July 2026, calling it an AI innovation hub for the nation.
Gouvernement de la République Gabonaise / Ministère des Nouvelles Technologies de l'Information et de la Communication — Gabon's Gouvernement, via the Ministère des Nouvelles Technologies de l'Information et de la Communication with UNESCO, established the Comité Technique National pour l'Intelligence Artificielle (CTN-IA) for AI governance in Gabon. The body coordinates national efforts toward an ethical AI deployment strategy governing the use of artificial intelligence-based technologies.
Ministry of Communication and Informatics (Kominfo), now Ministry of Communication and Digital (Komdigi) — The Minister of Communication and Informatics issued Ministerial Circular Number 9 of 2023 on the Ethics of Artificial Intelligence, the operative instrument for Indonesia AI regulation. Public and private business actors and electronic system providers must draft and apply in-house AI ethics policies built on inclusivity, humanity and security. As soft law, this AI governance tool carries no sanctions.
Ministry of Innovation, Science and Technology and Ministry of Justice — Israel's Ministry of Innovation, Science and Technology and Ministry of Justice built their Israel AI regulation strategy around one idea: "Responsible Innovation", set out in their policy document "Policy, Regulation and Ethics Principles in the Field of Artificial Intelligence." It lists six regulatory and six ethical principles matched to international standards, favouring sector-specific regulators, proportionate risk assessment, and non-binding guidance tools over hard law. This AI governance model keeps oversight decentralised across industries rather than concentrated in one central authority.
Ministry of Science, Technology, Knowledge and Innovation and Ministry Secretariat General of the Presidency — Chile now has non-binding AI in the public sector rules under Oficio Circular N° 711, issued jointly by the Ministry of Science, Technology, Knowledge and Innovation and the Ministry Secretariat General of the Presidency. State bodies building, buying or running artificial intelligence tools are urged toward human-centred design, explainable systems, stronger data protection, and staff training on government use of artificial intelligence. The rules carry no binding force on agencies.
Government of Spain, Ministry of Economic Affairs and Digital Transformation — Spain claims the European Union's first AI regulatory sandbox, created via Royal Decree 817/2023 (8 November 2023), where companies—chiefly SMEs and startups—pilot high-risk systems against draft EU AI Regulation demands ahead of enforcement. Working jointly with the European Commission, the project delivered 16 compliance guides through AESIA by December 2025, giving businesses practical AI compliance testing rather than theoretical guidance.
Parliament of Uruguay — Uruguay AI regulation stems from Article 74 of Law No. 20,212, which tasks Agesic with building a national artificial intelligence and data strategy spanning public bodies and private firms, guided by international benchmarks. The rule demands fairness, non-discrimination, accountability, transparency, auditability and cautious innovation, plus respect for human dignity and democratic order. Within 180 days Agesic owes lawmakers a report proposing rules balancing ethical progress, human rights and Uruguayan AI policy innovation.
Presidencia de la República Dominicana (Poder Ejecutivo) — Decree 498-23 (11 October 2023), Dominican Republic's Executive Branch, approved the National Artificial Intelligence Strategy (ENIA) — a Dominican Republic AI policy on four pillars: Smart Government, Talent and Innovation Hub, Data Hub, Regional Scale. Aims: automated citizen services, ethical governance, and regional leadership in artificial intelligence in government.
Datatilsynet (Danish Data Protection Authority) — Danish bodies building or rolling out AI systems must satisfy duties set by Datatilsynet: legal basis, informing those affected, and data protection impact assessments — new guidance on AI in public authorities. A companion survey tracked adoption across government nationwide.
Federal Public Service Policy and Support (BOSA) — Belgium's Raadgevend Comité Ethiek inzake Data en Artificiële Intelligentie (Advisory Committee on Ethics for Data and Artificial Intelligence), created by Royal Decree on 4 October 2023, shapes AI policy in government federally. Five experts on renewable five-year terms advise ministers on AI systems and data-driven technology, weighing rights, privacy, dignity and non-discrimination. This AI ethics in the public sector guidance is non-binding.
Executive Yuan — A 10-point framework now governs generative AI use by government agencies across Taiwan: the Executive Yuan cleared its Reference Guidelines for the Use of Generative AI by the Executive Yuan and Its Subordinate Agencies, binding from 3 October 2023. Human review must confirm every AI-assisted output before release, and staff are forbidden from feeding classified records, personal data or confidential business information into these tools. This public sector generative AI guidance places responsible conduct at the centre of agency practice.
Ministry of Communication, Telecommunications and Digital Economy (MCTEN) — Senegal AI policy rests on the National Strategy for Artificial Intelligence, unveiled 14 September 2023 by the Ministry of Communication, Telecommunications and Digital Economy: 'Stratégie nationale et feuille de route du Sénégal sur l'Intelligence Artificielle', to 2028. Four aims — the Plan Sénégal Émergent, AI for good, Ubuntu-inspired solidarity, trustworthy systems. 52 actions, six pillars: human capital, lab-to-market, an AI cluster, a regional AI hub for West Africa.
Saudi Data and Artificial Intelligence Authority (SDAIA) — The Saudi Data and Artificial Intelligence Authority (SDAIA) adopted the AI Ethics Principles, a Saudi Arabia AI policy covering every developer and user of AI systems. Its seven principles are accountability; transparency and explainability; reliability and safety; humanity; fairness; social and environmental benefits; and privacy and security. This national AI regulation uses a four-tier risk classification, with optional, penalty-free registration.
Jefatura de Gabinete de Ministros — The Jefatura de Gabinete de Ministros created the Interministerial AI Council (Mesa Interministerial sobre Inteligencia Artificial) via Administrative Decision 750/2023 to draft the executive branch's comprehensive national AI strategy. Chaired by the Secretariat of Innovation, Science and Technology, this Argentina AI policy body gathers ministries of Defence, Economy, Foreign Affairs, Health, Security, Human Capital, Justice, Deregulation and State Transformation, steering cross-sector AI coordination across economy and society.
Agencia de Acceso a la Información Pública (AAIP) — Argentina's AAIP created the National Programme for Transparency and Personal Data Protection in the Use of Artificial Intelligence via Resolution 161/2023, an AI governance Argentina framework for algorithmic accountability. It sets up an observatory and multidisciplinary advisory council, plus guides, training and campaigns on data protection and transparency in AI use.
Government of Spain, Council of Ministers — Spanish AI regulation gained a dedicated authority when Spain's Council of Ministers signed Royal Decree 729/2023, forming AESIA (Spanish Agency for the Supervision of Artificial Intelligence) under a Statute effective 3 September 2023. As designated EU AI Act enforcer, AESIA certifies systems, fosters ethical conduct, connects with European institutions, and watches deployments spanning government and industry, cementing artificial intelligence governance beneath one roof.
Conseil National de Transition (CNT) du Mali — Ordinance N°2023-022/PT-RM established the Centre d'Intelligence Artificielle et de Robotique du Mali (CIAR-Mali), a state body holding scientific, technical and cultural status and financed solely from national budget funds of roughly 3 billion FCFA. Under this Mali AI policy, the ordinance's 34 articles, grouped in five titles, charge CIAR-Mali with artificial intelligence and robotics research, development and training. The Council of Ministers endorsed the AI governance measure on 25 July 2023, ahead of unanimous ratification by the Conseil National de Transition on 19 October 2023.
Parlement de la République Gabonaise — The AI data protection law Gabon enacted defines artificial intelligence, along with raw data, AI input data and artificial neuron, making it the only statute in Francophone Afrique to spell out these terms. Loi n° 025/2023 revised the earlier Loi n° 001/2011 framework governing personal data processing. Article 77 bars courts and other bodies from basing a judicial ruling or any decision with legal effect solely on automated profiling of personality traits, while Article 43 grants individuals a right of access covering the reasoning behind such artificial intelligence regulation in personal data processing decisions.
Congreso de la República del Perú — Peru AI regulation now rests on Ley N° 31814, backing artificial intelligence uptake while putting human rights and individual welfare first. Required tenets: monitored safety, ethical design, privacy protection, transparent operation. Authority sits with the Presidencia del Consejo de Ministros, via its Secretaría de Gobierno y Transformación Digital, under this AI governance framework.
Ministry of Higher Education and Scientific Research; Ministry of Knowledge Economy, Start-ups and Micro-enterprises — Under Algeria AI policy, Minister Kamel Baddari's Ministry of Higher Education and Scientific Research joined Minister Yacine El-Mahdi Walid's Ministry of Knowledge Economy, Start-ups and Micro-enterprises to launch the Scientific Council for Artificial Intelligence (Conseil scientifique de l'intelligence artificielle), also known as the National AI Council. Professor Merouane Debbah leads 28 specialists from universities, business and public bodies. Tasks ahead: shape a national AI development strategy, gauge resources, refresh curricula, build partnerships.
Subsecretaría de Tecnologías de la Información, Jefatura de Gabinete de Ministros — Argentina's Subsecretariat of Information Technologies approved the 'Recommendations for a Trustworthy Artificial Intelligence', non-binding guidance for AI in the public sector, spanning government systems from design to retirement. As a trustworthy AI standard, not regulation, it lists seven principles: transparency, explainability, fairness, non-discrimination, privacy, oversight, accountability.
National Information Technology Development Agency — Nigeria's National Information Technology Development Agency (NITDA) drafted the Code of Practice for Artificial Intelligence, setting a nationwide baseline for AI systems. This Nigeria AI regulation mandates risk checks, disclosure, human oversight and conformity assessment for higher-risk systems, an emerging AI governance framework. Consultation continues as of September 2026, pending final adoption.
Cabinet of Rwanda — Cabinet approval made Rwanda the first African country with a comprehensive National Artificial Intelligence Policy, a landmark for Rwanda AI policy. MINICT and RURA built six pillars covering talent, digital infrastructure, data governance, government uptake, business expansion and ethical safeguards, targeting global standing in AI research and innovation. A dedicated AI Office inside MINICT now leads rollout, part of Rwanda's wider push on AI regulation.
National Council for Artificial Intelligence — The National Council for Artificial Intelligence endorsed the Egyptian Charter for Responsible AI at its seventh meeting, chaired by Communications and Information Technology Minister Amr Talaat. This Egypt AI policy mandates thorough impact assessments before government AI projects and gives ecosystem actors 13 general and 16 implementation guidelines. Five ethical principles anchor this national AI governance: fairness, accountability, human-centeredness, security and safety, transparency and explainability.
Ministry of Planning, Development and Special Initiatives — A 15-strong panel, the National Task Force on Artificial Intelligence, operates in Pakistan under Ahsan Iqbal to fast-track AI adoption in Pakistan across commerce, administration, schooling, medicine and growth targets. Charting a 10-year plan for these aims is its mandate. Ranks pair specialists with state and firm appointees, backing AI adoption in government.
Ministry of Research, Innovation and Digitalization — Ministerial ordinance OMCID 20484/2023 (Monitorul Oficial 382, 4 May 2023) sets the structure and operation of the Romanian Committee for Artificial Intelligence (CRIA). Government Memorandum 20D/31781/MN of November 2022 created CRIA as a Romania AI policy body under the Prime Minister's patronage. CRIA steers state and business AI projects, drafts rules and builds national AI governance capacity in research, development, education and deployment; MCID runs its technical secretariat.
Information Commissioner's Office — Fairness across the AI lifecycle sits at the centre of the ICO's revised guidance on AI and Data Protection, spelling out how UK GDPR duties reach accountability, transparency, lawfulness and accuracy in automated processing. The regulator sharpened its stance after industry pressed for clearer rules on AI data protection, backing safer adoption without weakening data protection in AI systems.
Government of Israel — Israel's National Program for Artificial Intelligence, adopted via Government Decision No. 173, commits up to NIS 500 million between 2023 and 2026 for research and AI infrastructure. A new coordination hub under the Ministry of Innovation, Science and Technology will handle AI policy implementation and rollout nationwide.
Ministry of Innovation, Science and Technology — Government Decision No. 173 tasked Israel's Ministry of Innovation, Science and Technology with building a central hub for coordinating national artificial intelligence oversight. This body, the Government Center for AI Regulation and Policy, links government agencies, advises on ethics and rulemaking, sets direction, handles cooperation abroad, and drives adoption of AI in public administration. Since September 2024 it has also chaired the national expert forum on the technology, anchoring Israel's approach to AI governance coordination.
Council of Ministers, Republic of Benin — Benin's Council of Ministers approved SNIAM, its national AI and big data strategy for 2023-2027, on Wednesday 18 January. Backed by 4.68 billion CFA francs, it bundles 123 actions across agriculture, education, health and public administration, forming the country's core AI policy framework.
Autoriteit Persoonsgegevens (AP) — Netherlands' Autoriteit Persoonsgegevens (AP) hosts the new Dutch Directorate for Coordination of Algorithms (DCA), tasked with AI supervision for systems that threaten fundamental rights. Rather than replacing existing regulators, the DCA supports them through cross-agency risk analysis, closer collaboration between supervisory authorities, and input into new legal frameworks for algorithm oversight.
Office of the Privacy Commissioner — New Zealand's Privacy Commissioner ties AI regulation in government agencies to the Privacy Act 2020's 13 Information Privacy Principles, covering a system's full life. Leaders sign off, assess privacy impact pre-launch, and keep someone accountable for outcomes. Vendor contracts must bar reuse of uploaded data for training, with Māori consulted on data sovereignty — AI governance for public sector agencies as an ongoing duty, not one step.
Palestinian Cabinet (Council of Ministers) — The Palestinian Cabinet approved the National Artificial Intelligence Strategy, grounding Palestine AI policy in a lasting AI industry fed by research, skills development and technological infrastructure. Five pillars structure the plan: AI for Government, AI for Development, AI for Capacity Building, AI and Legal Framework, and International Activities. The Ministry of Telecommunications and Digital Economy oversees delivery of this national AI strategy via the National AI Team.
Palestinian Cabinet of Ministers — A formal Palestine AI policy now exists: the National Artificial Intelligence Strategy, cleared by the Cabinet of Ministers and built by the Ministry of Telecommunications and Digital Economy to guide artificial intelligence responsibly and sustainably. Aims run from modernising administration to lifting outcomes in agriculture, water, healthcare and schooling, growing technical skills, and setting ethical and legal guardrails. This national AI strategy pairs growth with oversight.
Parliament of Niger (Assemblée Nationale) — Niger's Loi n° 2022-59, Article 52, sets AI regulation Niger relies on for automated decisions: judicial rulings cannot rest on automated processing assessing personality traits, and administrative or private-sector decisions cannot rest solely on automated profiling. Where a system qualifies as artificial intelligence, its criteria and personal data used must be disclosed at collection. President Mohamed Bazoum promulgated the statute, alongside Burkina Faso's as Africa's only explicit AI clause in data-protection law.
UAE Minister of State for Artificial Intelligence / AI Office — The Minister of State for Artificial Intelligence published the AI Ethics Principles and Guidelines, a non-binding UAE AI policy for designing, procuring and deploying AI across public and private sectors. Principles cover safety, privacy, fairness, transparency, accountability, human oversight and environmental sustainability. A self-assessment tool lets organisations test systems against this national AI governance approach.
Council of Ministers / Federal Public Service Policy and Support (BOSA) — The Belgian Council of Ministers adopted the National Convergence Plan for the Development of Artificial Intelligence on 28 October 2022, launching SmartAI.Nation. This Belgium AI policy rests on nine strands: trustworthy AI, cybersecurity, competitiveness, a data-driven economy, healthcare uses, sustainable mobility, environmental safeguards, lifelong learning and upgraded public services. It remains the federal blueprint for AI governance frameworks.
Ministry of Digital Economy and Entrepreneurship, approved by the Cabinet — A national roadmap now anchors Jordan AI policy: the Cabinet has signed off on the AI Strategy and Implementation Plan 2023-2027, aiming to position the kingdom among regional leaders in artificial intelligence. Priorities include growing homegrown expertise, funding research, attracting entrepreneurial investment and putting safeguards around responsible AI use. Officials also want AI deployed to sharpen efficiency across government and other priority industries, extending AI in public sector efficiency more broadly.
Government of the Republic of Tajikistan — The Government set Tajikistan AI policy through Government Resolution No. 483, the Strategy for the Development of Artificial Intelligence in the Republic of Tajikistan for the period up to 2040, targeting 5% of GDP from AI-related economic activity by 2040. The national AI plan prioritises ethical AI, model deployment, data enhancement and talent development. Resolution No. 83 amended it on 31 January 2025.
Ministry of Digital Economy and Entrepreneurship, approved by the Council of Ministers — Jordan's Council of Ministers signed off on the National Artificial Intelligence Code of Ethics, giving the Kingdom one moral baseline for AI systems, rooted in local religious and societal norms and matched to the UNESCO AI Ethics Recommendation. Jordan AI regulation now rests on seven named pillars: humanity/society, inclusiveness/justice, privacy/data, transparency, accountability/responsibility, reliability/integrity/authenticity, and environmental sustainability. This makes ethical AI development a formal goal of national policy.
Hellenic Parliament — Greece regulates artificial intelligence under Law 4961/2022, a national framework for emerging digital technologies and the country's core AI regulation Greece has enacted. Chapter B, Articles 3-14, mandates algorithmic impact assessments, transparency duties, and a public-sector registry of AI systems, alongside a Coordinating Committee for Artificial Intelligence overseeing this national AI framework.
Prime Minister's Cabinet Office / Ministry of Digital Economy and Society — Thailand's cabinet cleared a five-pillar national AI blueprint for 2022 to 2027: regulatory readiness, infrastructure, AI education, AI technology innovation, and sector adoption. This Thailand AI policy, under the Thailand National AI Strategy and Action Plan, targets an ecosystem lifting the economy and quality of life within five years, part of AI regulation in Thailand.
Council of Ministers, Federal Democratic Republic of Ethiopia — Council of Ministers Regulation No. 510/2022 creates the Ethiopian Artificial Intelligence Institute (EAII), an autonomous federal body reporting directly to the Prime Minister. This Ethiopia AI regulation assigns the EAII national AI research, policy advice, technology certification and national-security data analysis. That AI governance measure replaced the Artificial Intelligence Research and Development Center.
Electronic Transactions Development Agency (ETDA) — The Electronic Transactions Development Agency (ETDA) issued the Digital Thailand AI Ethics Guideline, a non-binding national framework advising regulators, developers, providers and users on keeping AI systems consistent with national strategy objectives, law and human rights. This ethical AI guidance rests on six core principles: reliability; security and privacy; transparency and accountability; competitiveness and sustainable development; equality, diversity, inclusiveness and fairness; and compliance with law, ethics and international standards.
National Science and Technology Development Agency (NSTDA) — Thailand's National Science and Technology Development Agency (NSTDA) runs a Thailand AI policy built on a non-binding standard, the NSTDA AI Ethics Guideline, steering research, development and deployment nationwide. Grant recipients, contractors, and public bodies conducting research must weave governance, privacy, transparency, fairness and safety through every stage of the AI lifecycle. The guideline also brings NSTDA's own research network into step with global AI governance norms.
Agência para a Modernização Administrativa (AMA) — Five evaluation pillars anchor GuIA Responsável (Responsible AI Guide), the framework Portugal's Agency for Administrative Modernisation (AMA) built for AI in public administration: accountability, transparency, explainability, fairness and ethics. A bundled scoring instrument lets teams assess systems against those pillars ahead of rollout. AMA aimed the guide at government bodies, yet academic institutions and private firms may equally adopt this AI governance framework for the public sector, since its scope is not confined to state agencies. Its debut coincided with the launch of AI4PA.
Instituto Nacional de Transparencia, Acceso a la Información y Protección de Datos Personales — Mexico AI regulation guidance from the National Institute for Transparency, Access to Information and Personal Data Protection (INAI) covers personal data tied to artificial intelligence use across government bodies and companies. It sorts AI into categories, sets principles for accountable AI governance, and gives controllers a checklist for confirming compliance during AI product development. The move reflects Mexico's wider AI data protection policy.
Ministry of Science, Technology and Innovation (MOSTI) — Under national AI policy, Malaysia's MOSTI rolled out the National Artificial Intelligence Roadmap 2021-2025 (AI-Rmap), a five-year blueprint for a lasting domestic AI innovation base. Thrusts span governance, R&D, infrastructure, skills and uptake, with 11 flagship applications across five fields: farming and forestry, health, smart urban mobility, learning, and government services. Seven values anchor trustworthy AI deployment: fairness, dependability, privacy, inclusion, openness, accountability and benefit to people.
Department for Digital, Culture, Media and Sport (now Department for Science, Innovation and Technology) — This UK AI policy, the National AI Strategy, sets a decade-long plan to expand Britain's AI sector, spread adoption, and frame pro-innovation UK AI regulation. It unites funding, skills, oversight and infrastructure to build the UK as a top global AI power.
Cabinet of Qatar / Ministry of Communications and Information Technology — Qatar AI policy took shape through Cabinet Decision No. (10) of 2021, which sets up a National Artificial Intelligence Committee based at the Ministry of Communications and Information Technology (MCIT), tasked with rolling out and tracking progress on the country's National AI Strategy. The inter-ministerial body coordinates AI-related work across government entities, proposes programmes to build human capacity, backs AI start-ups, and speaks for Qatar at international AI forums. This reflects Qatar AI regulation extending into national AI governance and oversight; the decision was issued 1 September 2021.
Palestinian Council of Ministers — Palestine's National AI Team, formed 6 September 2021 by the Council of Ministers under the Ministry of Telecommunications and Digital Economy, seats ministries, academics and private-sector figures. Its brief: vet AI policy, feed the national AI strategy, order priorities, steer rollout follow-up, build partnerships and handle AI ethics — an AI governance structure covering both direction and oversight.
Swiss Federal Council — Switzerland's Federal Administration centralised AI in government via the Competence Network for Artificial Intelligence (CNAI/KNW KI), created by Swiss Federal Council decision, first hosted at the Federal Statistical Office. One unit covers advice, tooling and cross-agency coordination on government AI adoption, ending fragmented expertise. Six hubs split duties: governance and strategy, platforms, procurement, culture and training, law and regulation, cyber security, and data science. Since 1 February 2026 the Federal Chancellery, not the statistics office, runs the contact point.
Ministry of Transport, Communications and Information Technology — Oman's government ministries and state units gained fresh direction on AI adoption in government through a Ministry of Transport, Communications and Information Technology circular, built on six values: inclusiveness, empathy, accountability, equity, transparency, security. Officials frame public-sector artificial intelligence use as advancing Oman Vision 2040 and Fourth Industrial Revolution goals.
Department of Enterprise, Tourism and Employment — Under Ireland AI policy, the Department of Enterprise, Tourism and Employment unveiled 'AI - Here for Good', the nation's inaugural National AI Strategy, targeting a top global position in trustworthy, human-centred artificial intelligence. Officials back an ethical, people-focused path for building, deploying and applying AI, meant to boost output, tackle social problems and upgrade public services via responsible AI adoption.
Ministério da Ciência, Tecnologia e Inovação (MCTI) — Brazil set its AI policy in Brazil through the Estrategia Brasileira de Inteligencia Artificial (EBIA), created by Portaria MCTI no 4.617 and revised by Portaria no 4.979 on 13 July 2021. Grounded in OECD ideals — fairness in growth, focus on people, openness, resilience and answerability — the Strategy shapes state pursuit of research, innovation and responsible artificial intelligence deployment. Nine pillars organise the plan: law, governance, foreign relations, skills training, R&D, and rollout across industry, government and public safety.
Digital Policy Office — The Digital Policy Office's Ethical AI Framework, anchoring Hong Kong AI policy, steers government bureaux and departments to design and deploy ethical AI and big data analytics IT projects. Revised August 2023 and December 2025 for generative AI, this ethical AI guidance bundles guiding principles, leading practices and an AI assessment template.
Austrian Federal Government (Federal Chancellery and Federal Ministry for Innovation, Mobility and Infrastructure) — Austria's national AI strategy, Artificial Intelligence Mission Austria 2030 (AIM AT 2030), rests on three pillars: trustworthy adoption rooted in European fundamental values, an AI research and innovation hub, and economic competitiveness. Health, climate, industry, education and public administration fall within scope. The document is Austria's central AI policy framework.
Government of Iceland (Prime Minister's Office) — Iceland's Prime Minister's Office built its Iceland AI policy, formally the national Policy on Artificial Intelligence, Stefna Íslands um gervigreind, on three pillars: benefit for everyone, a competitive workforce, and schooling fit for modern times. The plan binds government to ethical technology, adaptable jobs and protection of the Icelandic language online. It anchors later governance work, including the AI Action Plan 2025-2027, as the country's founding national AI governance framework.
Government of Israel — Israel handed its Minister of Innovation, Science and Technology overall charge of AI policy Israel-wide under Government Resolution No. 212, spanning regulation, data rules, ethics, cross-border cooperation and public-sector uptake. An inter-ministerial team, chaired by the Ministry's Director General, now studies artificial intelligence questions and drafts proposals. The setup concentrates national AI governance under one office rather than scattering it across departments.
Ministry of Digital Economy and Entrepreneurship, approved by the Council of Ministers — Jordan's Council of Ministers fixed the country's Jordan AI policy, adopted 20 December 2020, covering every major part of the economy and readying legal, regulatory and technical foundations for wider use of this Jordanian artificial intelligence strategy. Five strands hold the plan together: oversight structures, rules and legislation, digital networks, conditions for business and investment, and skills-building. Implementation sits with a new National Ministerial Committee for Artificial Intelligence, led by the Minister of Digital Economy and Entrepreneurship.
Council of Ministers of the Republic of Bulgaria — Cabinet ministers backed the Concept for the Development of Artificial Intelligence in Bulgaria until 2030 via Protocol No. 72. Six linked strands make up this Bulgaria AI policy: scientific infrastructure, communications, lifelong learning, education, tech transfer, research, data rules, sector-based innovation, and an ethics framework matched to EU trustworthy-AI norms — an artificial intelligence strategy running to 31 December 2030.
Government of Spain, Ministry of Economic Affairs and Digital Transformation — Spain adopted ENIA, its National Artificial Intelligence Strategy, directing public funds toward AI in public administration and industry between 2021 and 2023. Sitting under Component 16 of the Recovery, Transformation and Resilience Plan, the scheme forms part of the broader España Digital 2025 push. It rests on six pillars: strengthening research, growing digital talent, scaling data infrastructure, embedding AI within value chains, extending it into government agencies, and establishing an ethical and regulatory framework. These figures come from the Spanish government's own stated commitment, not an independently audited result.
Agencia de Gobierno Electrónico y Sociedad de la Información y del Conocimiento (Agesic) — Agesic's Guide for Algorithmic Impact Study (EIA) gives public-sector teams a structured algorithmic impact assessment method for automated decision-support systems built on machine learning. It helps project leads flag key risks and design considerations before deployment. The guide supports Uruguay's broader approach to AI governance impact assessment in government projects.
Ministry of Investments, Regional Development and Informatization of the Slovak Republic (MIRRI) — MIRRI created CERAI, the Permanent Commission for Ethics and Regulation of AI, as an independent expert advisory body on Slovakia AI regulation. The commission's 21 members weigh ethical, social and legal questions in regulating artificial intelligence across research, development, deployment and use, under a February 2021 statute.
Government of the Russian Federation — Government Directive No. 2129-r introduced Russia's Concept for the Development of Regulation of Relations in the Field of Artificial Intelligence and Robotics until 2024, charting the state's path on AI regulation and robotics policy. Officials laid out objectives, guiding principles and sector-by-sector duties spanning accountability rules, data handling regimes, technical benchmarks and insurance, alongside deploying AI across hospitals, transport networks, banking, smart urban systems and government agencies. This foundational text for AI governance rules continues in force under the Russian Federation's legal system.
Agency for the Assessment and Application of Technology (BPPT), now part of the National Research and Innovation Agency (BRIN) — BPPT built Stranas KA (Strategi Nasional Kecerdasan Artifisial), Indonesia's 2020-to-2045 blueprint fixing Indonesia AI policy on artificial intelligence. Priority falls on five fields: medicine, government reform, schooling with research, food supply, plus transport and smart cities. Ministries, regional offices, universities and firms answer to one artificial intelligence governance roadmap spanning that quarter-century.
Saudi Data and Artificial Intelligence Authority (SDAIA) — Saudi Arabia AI strategy plans, formally the National Strategy for Data and AI, set 2030 as the target year for ranking among the world's top data-driven economies. The Saudi Data and AI Authority (SDAIA) drew up the roadmap, later signed off by King Salman bin Abdulaziz Al Saud, mapping out phases that move from addressing urgent domestic needs, to strengthening long-term competitiveness, to becoming a major global supplier and consumer of data and artificial intelligence.
President of the Russian Federation — Moscow gained a dedicated legal sandbox for testing artificial intelligence under Federal Law No. 123-FZ, which President Putin put into force to build conditions favourable for developing and rolling out AI systems within the Russian capital. Anonymised personal data may now be processed specifically to advance these AI regulation Moscow goals, alongside rules spelling out the trial's aims, guiding principles and operating procedure. Russia's State Duma cleared the measure on 14 April 2020, and the Federation Council gave final sign-off three days later, on 17 April 2020, completing the artificial intelligence experimental legal regime's legislative path.
Spanish Data Protection Agency (AEPD) — The Spanish Data Protection Agency (AEPD) issued a guide on Spain AI regulation titled 'Una aproximación para la adecuación al RGPD de tratamientos que incorporan Inteligencia Artificial' (An approach to GDPR compliance for processing that incorporates Artificial Intelligence). The guide covers privacy risks when regulating AI, plus key GDPR points for AI products and services: transparency, lawfulness of processing, risk management and automated decision-making.
Cabinet of Ministers of Latvia — Latvia's Cabinet of Ministers approved a national AI strategy via the report 'Par mākslīgā intelekta risinājumu attīstību' (Developing Artificial Intelligence Solutions), covering six areas: skills, trial projects, language technology, ethics, infrastructure and innovation. This AI adoption roadmap guides public and private uptake of artificial intelligence economy-wide.
Council of Ministers — Cyprus's Council of Ministers adopted the National Artificial Intelligence Strategy of Cyprus, drafted by the Department of Electronic Communications and an AI panel. Four pillars anchor it: partnership investment, data spaces, lifelong learning, and trustworthy AI, forming the basis of Cyprus AI policy.
Information and Communication Technology Division (ICT Division) — Six pillars anchor Bangladesh AI policy: agriculture, health, education, manufacturing and public service delivery, defined by the Information and Communication Technology (ICT) Division through the National Strategy for Artificial Intelligence. Execution spans a five-year national AI strategy timeline to 2024, with a proposed task force and steering committee reporting to the Division.
Ministry of Science and ICT — MSIT built South Korea's national AI policy on 100 tasks across 9 strategies spanning AI ecosystem, AI utilisation, and people-centred AI. Cabinet adopted this as the National Strategy for Artificial Intelligence, slogan 'Toward AI World Leader beyond IT', targeting competitiveness and living standards by 2030 — marking AI adoption in South Korea as a cross-ministry drive.
Secretaría de Ciencia, Tecnología e Innovación Productiva, Presidencia de la Nación — Argentina's Plan Nacional de Inteligencia Artificial, known as ArgenIA, rests on six pillars: talent, data, computing infrastructure, research, sectoral uptake and ethics. The Secretariat for Science, Technology and Productive Innovation backs the roadmap to 2030, framing national AI policy and public-sector AI adoption across government and business.
Cabinet of Egypt, Prime Minister — Egypt's National Council for Artificial Intelligence traces its authority to a 2019 decree from Mostafa Madbouly, sitting within the Prime Minister's office and led by the Minister of Communications and Information Technology. As the leading force behind Egypt AI policy, it drafts the National AI Strategy alongside ethics guidance for wider AI governance.
President of the Russian Federation — President Vladimir Putin's Decree No. 490 launched Russia's National Strategy for the Development of Artificial Intelligence — a Russia AI strategy stretching to 2030 that covers research capacity, computing infrastructure, data access, skilled talent and uptake throughout government and industry. Decree No. 124 revised the plan on 15 February 2024, introducing fresh benchmarks for home-grown computing power, large-scale generative models, how widely organisations deploy AI, and trust rules governing public-sector systems. This places Russian national AI policy on a longer horizon than most comparable national plans.
Dutch government (Ministry of Economic Affairs and Climate Policy) — SAPAI, the Dutch government's Strategic Action Plan for Artificial Intelligence, sets over a hundred measures pairing public investment with joint work between industry and researchers. It funds entrepreneurs and scientists, builds digital infrastructure and data systems, and adds ethical and legal safeguards for citizens' rights under this national AI plan. The AI strategy targets economic and societal gains from AI for the country.
Malta Digital Innovation Authority and the Maltese Government — Malta now has a national AI strategy: 'The Ultimate AI Launchpad: A Strategy and Vision for Artificial Intelligence in Malta 2030', from the Maltese Government's Malta.AI Taskforce under the Malta Digital Innovation Authority. Three pillars anchor it — investment, public and private sector uptake — backed by education, ethics and infrastructure, plus 72 actions. Six pilots span traffic, health, tourism, utilities and other services, aiming for a top-ten global rank by 2030 — Malta AI policy in practice.
Ministry of Science and Technology (now NSTC) — Eight ethical principles anchor Taiwan AI technology regulation under the AI Technology R&D Guidelines from the Ministry of Science and Technology, built on three core values: human-centred development, sustainability, and diversity and inclusion. Fairness, autonomy, safety, privacy, transparency, explainability, accountability and the common good together form the basis of the country's AI research and development governance.
Custodian of the Two Holy Mosques (Royal Court of Saudi Arabia) — A royal decree in Saudi Arabia set up SDAIA, the Saudi Data and Artificial Intelligence Authority, as the kingdom's single national body for AI regulation and data policy under Royal Order No. (A/471). Reporting straight to the Prime Minister, SDAIA governs three arms: the National Data Management Office, the National Center for AI, and the National Information Center, forming the backbone of Saudi data and AI governance.
Government of the Czech Republic — The Committee for Artificial Intelligence (Výbor pro umělou inteligenci) oversees delivery of Czechia AI policy under the first National AI Strategy, tracking targets and endorsing action plans. Housed in the Government Council for the Information Society, within the Ministry of Industry and Trade, this AI governance body gathers several ministries, the Technology Agency, the Academy of Sciences and other stakeholders.
Ministry of Industry, Business and Financial Affairs (Erhvervsministeriet) — Four goals anchor Denmark's national artificial intelligence strategy: putting people first in AI ethics, building up Danish research capacity, helping companies expand through AI adoption, and pushing public services toward world-class delivery. Six ethical principles and 24 initiatives back the plan, funded with DKK 60 million across 2019 to 2027.
Malta Digital Innovation Authority — Malta AI regulation rests on 'Towards Ethical and Trustworthy AI: Malta's Ethical AI Framework', from the Malta Digital Innovation Authority. Four pillars — autonomy, harm prevention, fairness, explicability — pair with 63 controls for trustworthy AI governance. It draws on EU High-Level Expert Group guidelines and OECD AI Principles, tying to Malta's national AI strategy, with oversight under a National Technology Ethics Committee at the MDIA.
Ministry of Transport and Communications (now Ministry of Communications and Information Technology) — Qatar AI policy rests on the National Artificial Intelligence Strategy, signed at QITCOM, which structures government action around six themes: schooling, data sharing, jobs, enterprise, scholarship and ethical safeguards. Officials cast this as an 'AI+X' paradigm intended to spread machine-learning use through every part of the economy. The national AI strategy still guides Qatar into 2026, tracked by the Artificial Intelligence Committee for compliance.
Federal Government of Germany (Bundesregierung / BMFTR) — German AI policy took formal shape when the Federal Cabinet approved the Federal Government's Artificial Intelligence Strategy, aiming to boost Germany's standing in AI research, development and application against international rivals. Revisions in December 2020 brought pandemic response and sustainability into scope, and the BMFTR later issued its AI Action Plan, mapping eleven fields of action and committing more than 1.6 billion euros in funding. This sequence marks Germany's shift from an initial artificial intelligence strategy toward binding investment pledges.
NITI Aayog — NITI Aayog released the National Strategy for Artificial Intelligence (#AIforAll), setting India AI policy under the Finance Minister's mandate. Priority areas span healthcare through agriculture, education, smart cities and infrastructure, plus smart mobility and transportation, followed by a two-tiered research structure — Centres of Research Excellence alongside International Centres of Transformational AI — as its artificial intelligence adoption framework.
Gouvernement français — France's national AI strategy, the Stratégie nationale pour l'intelligence artificielle (SNIA), runs across two phases, 2018-2022 and 2021-2025, targeting global and European leadership in artificial intelligence. It covers all French territory, including French Guiana, focusing on research strength, talent training, economic AI uptake, and ethics. COSNIA, within the Direction générale des Entreprises, coordinates the artificial intelligence development plan.
French Government (Presidency of the Republic) — French AI policy traces to the Stratégie nationale pour l'intelligence artificielle, unveiled by Emmanuel Macron at the Collège de France during the AI for Humanity colloquium. Four pillars structure it: talent and research for an AI ecosystem, open data for domestic champions, funding for faster AI adoption in health and mobility, and an ethical, political framework. Mayotte, a French overseas department, falls under this Strategy like the rest of France.
UAE Cabinet — The UAE Cabinet formed the UAE Council for Artificial Intelligence, chaired by the Minister of State for Artificial Intelligence, to steer AI adoption across government departments and education. This is UAE AI policy in practice: rules for an AI-friendly ecosystem, advanced research, and cooperation between federal and local entities under national AI governance.
Government of France — France launched its national AI strategy, the Stratégie nationale pour l'intelligence artificielle (SNIA), unveiled by President Emmanuel Macron at the Collège de France and building on the Villani report 'Donner un sens à l'intelligence artificielle'. It targets European and global AI leadership through two funded phases: Phase 1 (2018–2022, €1.5 billion) for research capacity, Phase 2 (2021–2025, €1 billion) for spreading AI use economy-wide. Martinique falls under this artificial intelligence policy as French territory.
President of the Republic of Belarus — AI development Belarus gained legal grounding when the President's Decree No 8 took effect on 28 March 2018, licensing artificial intelligence, machine learning and neural network systems within the High Technology Park (HTP), plus autonomous vehicles and blockchain. HTP's tax and legal terms were extended to 2049. BelTA reports resident expertise now spans artificial intelligence in Belarus's digital economy, machine learning, Internet of Things and augmented reality — the root document centring AI in national digital strategy.
Ministry of Economic Affairs and Employment — Finland AI strategy, issued by the Ministry of Economic Affairs and Employment, aims to make the country a global leader in applying artificial intelligence across business, government services and society. Titled 'Finland's age of artificial intelligence', it sets out recommended actions for the Nordic nation to capitalise on AI-driven growth. The Finnish AI policy framework targets adoption across multiple sectors rather than one industry.
UAE Cabinet — UAE National Strategy, adopted by the Cabinet, targets worldwide dominance in artificial intelligence by 2031. UAE AI strategy spans eight pillars touching healthcare, transport, energy and schooling: cultivating an AI destination, nurturing a fertile ecosystem, and weaving AI adoption in public services into everyday dealings between residents and the state. Officials brand it pioneering, unmatched elsewhere across the region or the planet.
Kanazawa Medical University Hospital — Kanazawa Medical University Hospital went live with AI electronic medical records, deploying Ubie's clinical documentation AI after a July trial. Chosen over two rivals for lower cost and reach across roughly 1,800 hospitals nationwide, the platform aims to cut physicians' paperwork to a tenth.
Mie University Hospital — Apollo AI, an AI core system from Albatrus, now runs inside Mie University Hospital, tied into its electronic medical records under the MUDX plan. Staff across departments use built-in agents to draft documentation and surface patient information, reflecting Japan's shift toward AI-assisted medical records. A cross-professional group was formed to steer uptake.
AdventHealth — AdventHealth opened OpenEvidence, an AI clinical evidence search tool, to therapists, pharmacists, nurses, physicians, care managers and other clinical roles. Evidence lookup sits inside the electronic health record.
Atyrau Region health authorities (Kazakhstan) — Fifteen polyclinics and primary-care organisations across Atyrau Region adopted Akylmed AI, handing AI appointment booking to Asel, a round-the-clock virtual operator. City Polyclinic No. 6 already covers over 37,000 patients through this AI patient contact system.
Tanner Health — Tanner Health became the first commercial US user of AI ECG heart attack detection, running Powerful Medical's PMcardio platform with the Queen of Hearts model after FDA De Novo authorisation. Suspected heart attack patients across west Georgia and east Alabama get AI-read ECGs in a 911 ambulance, community hospital or cardiac centre.
Bulgarian Cluster for Digital Solutions and Innovation in Healthcare — The Bulgarian Cluster for Digital Solutions and Innovation in Healthcare is touring 29 settlements with a free mobile AI eye screening unit, alongside eye-care and pharmacoeconomics associations, Roche Bulgaria and the Ministry of Health. Software checks photos for early age-related macular degeneration, diabetic retinopathy or glaucoma. AI-assisted sight testing runs 28 September to 13 November across six provinces, Sofia, Pleven, Varna, Burgas, Stara Zagora and Plovdiv, plus a fixed Sofia site.
University of Technology and Applied Sciences — SafeSiso delivers confidential reproductive guidance to Ghanaians aged 10 to 19, an AI chatbot for adolescent reproductive health built by two University of Technology and Applied Sciences students on WhatsApp. Launched in Navrongo, it now runs across Upper East Region schools with a referral link to health professionals. Girls needing judgement-free health answers form the core audience for this adolescent health chatbot.
Yaroslavl Region Ministry of Health — Yaroslavl region's health ministry will put AI in cardiology consultations by late 2026, acting deputy health minister Svetlana Belyakova told a regional Duma committee in Russia. Planned functions: sorting circulatory-disease case data, compiling call-in rosters, flagging comorbid patients, recommending therapy adjustments or extra tests, and scoring check-up risks. This clinical decision support starts with heart care.
HEALWELL AI — HEALWELL AI opened SMART Summary to more than 400 New Brunswick healthcare providers, producing AI clinical summaries inside Intrahealth's Profile electronic medical record within seconds. Of 50 clinicians piloting this AI chart review tool across Ontario, British Columbia and New Brunswick, 84% foresaw time savings and 79% would recommend adoption.
Ministry of Health (Côte d'Ivoire) — Côte d'Ivoire's Ministry of Health will widen AI health logistics from one Zipline hub at Daloa to ten, following government sign-off on an eight-year public-private partnership. Planners expect 2,200-plus clinics and upwards of 12 million residents to be reached once complete. The US State Department puts up US$150 million, paid for performance, towards construction; the treasury meets running costs of this drone-based medicine distribution.
Prefeitura de Fortaleza — Fortaleza is tracking roughly 5,000 residents with hypertension or diabetes at three municipal clinics through IAra, a WhatsApp AI care assistant inside the Saúde que Cuida Digital bundle for the SUS. Enrolment goals sit at 20,000 by the end of 2026 and 200,000 by the end of 2027 for this chronic-disease follow-up tool. Alongside came the Saúde na Palma da Mão app plus booking confirmation over WhatsApp.
US Department of Veterans Affairs — The Department of Veterans Affairs has opened Abridge's AI scribe to clinicians at every VA medical centre after a 75-centre pilot spanning the new Oracle EHR and legacy VistA/CPRS. Beyond transcribing and drafting notes, the documentation assistant briefs doctors pre-visit, prompts talking points mid-appointment and proposes follow-ups and orders.
Ministry of Health; Centrum e-Zdrowia — Poland's Ministry of Health will open national AI in radiology to hospitals before year-end via the Platforma Usług Inteligentnych (PUI), which Centrum e-Zdrowia pitches as a digital assistant for radiologists. First come AI image analysis aids for X-ray, CT and MRI scans, paid from the e-Health strand of the National Recovery and Resilience Plan.
Kings Hospital Colombo — Kings Hospital Colombo has opened what the hospital calls Sri Lanka's earliest hospital AI chatbot, fielding queries in ordinary speech at any hour. The patient-facing assistant covers appointments, consultants, channelling and checkups.
Regione Umbria — Regione Umbria, with Usl Umbria 1 and Usl Umbria 2, has put Mia before its GPs: an Agenas-backed system bringing AI for family doctors under mission 6 'Salute' of the Piano Nazionale di Ripresa e Resilienza. Around 60 sat through the Broletto workshop, Perugia, on 16 September. Mia throws up guidance, treatment routes and appropriateness rules, leaving the call and rapport with the patient to the clinician — AI in general practice that defers.
Chi Mei Medical Center — Chi Mei Medical Center now gives A&E doctors AI medical history summaries, putting past illness and prescriptions on one page in 30 seconds. An on-premises language model builds these automated patient summaries, which once took five to ten minutes to compile.
Ministry of Health (Greece) — Health Minister Adonis Georgiadis launched MedWatch on Friday 18 September, setting AI in medical regulation on what doctors and clinics say about themselves online. The tool trawls medical association registries, Google Maps and Google Places, clinic sites, adverts on Meta, TikTok and Google and public profiles, then weighs every claim against the named practitioner's specialty, licence and lawful setting. Athens surgeries fell under the healthcare AI sweep straight away, the rest of Greece inside roughly 10 days.
Taipei City Hospital — Taipei City Hospital has adopted AI ECG interpretation so clinicians on the front line catch probable acute myocardial infarction, dangerous arrhythmias and similar red flags. Launched alongside a western critical-care heart unit at Zhongxing, connected with Renai, the heart-trace screening also supports firefighters developing warnings for cardiac emergencies.
Nagai Hospital — Nagai Hospital in Mie Prefecture has pushed staff adoption of an AI documentation deployment to roughly 95% after installing Albatrus's Apollo AI platform, now the backbone of its NAGAI 100 centenary DX push and wired directly into the MI·RA·Is electronic medical record infrastructure. Drafting a discharge summary dropped from 37 minutes to 3, informed-consent transcription shrank from 15 minutes to 4, spanning discharge summaries, nursing notes and symptom records. Few disclosed cases of EMR-linked clinical note automation in Japan show gains this sharp.
Canadian Mental Health Association Waterloo Wellington — Canadian Mental Health Association Waterloo Wellington is the first organisation running VitalHub AI Scribe, an AI scribe for mental health casework in the CaseWORKS case management solution. The ambient tool listens, extracts clinically relevant information and writes the paperwork: transcriptions, case notes, client and patient summaries. Chief executive Helen Fishburn says growing documentation requirements weigh on staff, and that AI documentation frees team members for clients and community.
Baystate Health — Baystate Health has begun AI weapons detection, starting Monday at Baystate Wing Hospital in Palmer, where visitors are scanned on arrival without holding up foot traffic, while phones, keys and wallets still pass. Public safety director Jeff Hauk says the hardware and the AI platform carry out the screening, with other ways to check anyone drawing higher scepticism. Next the same screening technology goes to the emergency department at Baystate Medical Center for a test run. Deployment at the rest of Baystate's community hospitals follows through the autumn, and phase two would reach main entrances.
Unidade de Pronto Atendimento 24h de Rondonópolis — Unidade de Pronto Atendimento 24h de Rondonópolis has begun scanning with a machine carrying AI ultrasound diagnostics, bought by the municipality for about R$278,000 of its own money. Onboard intelligence shortens waits for results in severe and emergency presentations, and echocardiography joins the unit's procedure list for the first time. AI-assisted imaging here has had no published independent evaluation.
Altai Krai Medical Information and Analytics Centre (AKMIATs) — Russia's Altai Krai medical information and analytics centre will install AI prescribing support for doctors under a 15.6 million rouble tender. Hooked up to the territory's healthcare information system, the software will screen orders, catch contraindications and harmful combinations and score side-effect risk. Bidding closes 22 September, delivery by 1 December 2026, and the Moscow region adopted comparable clinical decision support in February 2026.
Loblaw Companies (PC Health) — Loblaw's PC Health has launched PC Chat, a free AI health tool answering questions in plain language from Canadian clinical guidelines and 1,000-plus Canadian clinical sources. It runs in most provinces on the PC Health and Shoppers Drug Mart apps and sites; New Brunswick and Quebec follow this autumn. Medical professionals on the Clinical AI Advisory Committee helped build it, and it suggests next steps such as seeing a healthcare provider.
Sociedade Médica Cabo-verdiana Americana (CVAMS) — Sociedade Médica Cabo-verdiana Americana, CVAMS, has run the Continuidade project in Cabo Verde for six years, and AI diagnosis in Ophthalmology now enables nurses, not doctors, to pin down cataracts, glaucoma and retinopathy. Cloud-linked machines generate automatic diagnoses, aimed at Oncology and Cardiology, the specialities where local capacity is thinnest, under surgeon Júlio Tavares. The current mission, 10 to 24 September, fields over 70 doctors, surgeons, nurses, psychologists and further health workers drawn from leading United States institutions. CVAMS casts its Artificial Intelligence diagnostic equipment as a weapon against the archipelago's foremost killers; no independent evaluation of accuracy has been published.
Comunidad de Madrid — Comunidad de Madrid will use AI in chronic disease monitoring for close to 2 million patients through CRON-IA, a system flagging early warning signs in their records. Personaliz@ extends remote monitoring for hospital patients recovering at home via continuous assessment and clinical alerts, while PREVEN-IA builds digital twins from genomic, analytical, imaging and clinical data to forecast outcomes, an AI-based patient monitoring approach across three tools. Ayuso outlined the plan at the Madrid Assembly; none of the systems is installed yet, and these remain commitments rather than deployments.
Clinical Center of Montenegro (KCCG) — The Clinical Center of Montenegro (KCCG) and IT Advanced Services (ITAS) built PathAI with a nearly 193,400-euro Innovation Fund of Montenegro grant, applying AI in pathology to digitised slides and helping pathologists recognise tumour tissue. PathAI's two-stage architecture detects tumours, then segments multi-class tissue; AI-assisted histopathology validation began with colorectal adenocarcinoma.
Clinical Center of Montenegro (KCCG) and IT Advanced Services (ITAS) — Montenegro clinicians now get machine backup for spotting cancer in tissue slides: KCCG teamed with ITAS to build PathAI, software applying AI tumour diagnosis pathology to digitised histopathological samples. Nearly 200,000 euros from the Innovation Fund of Montenegro funded the pilot phase, aimed at colorectal adenocarcinoma. Final calls stay human — pathologists, not the cancer diagnosis software, make the diagnosis.
Insilico Medicine — Insilico Medicine's rentosertib, whose target and molecule both came from generative AI and which reported a Phase 2a trial in idiopathic pulmonary fibrosis (Nature Medicine, June 2025), has started its Phase III trial. The randomised, double-blind, placebo-controlled study (NCT07687459), registered in July 2026, enrolled its first patient on 9 September and aims to enrol 320 patients over 52 weeks.
Moscow city health service — Moscow has given all district doctors in its adult polyclinics an AI note-taker, a voice assistant that fills in consultation records for more than 3,000 clinicians. Running in the background, it transcribes doctor–patient conversation and logs diagnoses, complaints, history and recommendations in the protocol, which the doctor reviews and confirms before signing. Deputy mayor Anastasia Rakova said earlier tests with neurologists more than doubled doctors' time with patients, and experts rated AI-written protocols 4.8 out of 5.
Proximie and eight NHS hospital sites — Proximie's Intelligence Suite, a surgical AI platform, is going into 104 theatres, endoscopy suites and catheterisation labs at eight NHS sites in the United Kingdom, under a year-long evaluation with Amazon Web Services and Deloitte. Participating trusts, including University Hospitals Coventry and Warwickshire and Barking, Havering and Redbridge University Hospitals NHS Trust, together handle 20,000-plus procedures yearly through this hospital operating theatre AI programme. The aim: shrink gaps between cases, not lengthen shifts. It records proceedings without interrupting clinicians, strips personal details automatically, and turns footage into data flagging where lists stall.
Adventist Health — Adventist Health finished moving 27 hospitals and 440-plus clinics in California, Oregon and Hawaii to Epic on 30 August, launching what it called the largest AI-enabled tool deployment in a single Epic go-live. It says it now uses more of Epic's AI-powered tools at launch than any US health system before.
TABIB — TABIB has AI radiology pilots running now in Azerbaijan, alongside tools aimed at easing doctors' workflows. The state health body ranks machine reading of medical images, earlier disease detection, clinical decision support and patient-data handling among the more promising uses of medical imaging AI. Candidate systems are still being weighed, and the public will hear which clear the bars of clinical accuracy, data security, personal-data protection and specialist supervision.
Benin National Malaria Control Programme — Benin gave South Korean diagnostics supplier Noul a public tender for 20 AI malaria diagnosis analysers for severe-malaria hospitals, and the National Malaria Control Programme has endorsed the miLab MAL platform as a point-of-care device. The system handles each step from blood-smear preparation to digital imaging and AI analysis, making AI-assisted testing quicker and more consistent, Noul says. The purchase followed a 2025 clinical evaluation of 211 children with suspected severe malaria at northern and southern university hospitals.
Nemocnica Malacky — Nemocnica Malacky's refurbished suite now holds Slovakia's first Philips Ingenia Ambition X, using Dual AI software for AI MRI scanning. The 1.4-million-euro system speeds AI-assisted magnetic resonance imaging and sharpens pictures, rebuilding frames and correcting breathing motion. Its 70-centimetre bore gives patients more room.
Chinese PLA General Hospital (301 Hospital) — The Senior Department of Cardiology at the Sixth Medical Center of Chinese PLA General Hospital, better known as 301 Hospital, completed a transcatheter closure for congenital heart disease on a 48-year-old man with an AI-powered ultrasound robot at the probe. The robot carried out image acquisition, lesion localisation and real-time navigation on its own through the procedure, and the team leaned on it to choose the occluder and place it accurately. The hospital calls this a first for AI in interventional cardiology, resting that on a search it attributes to an authoritative institution.
Hospital Dr. Rafael Ángel Calderón Guardia (CCSS) — Hospital Dr. Rafael Ángel Calderón Guardia, run by Costa Rica's social security fund (CCSS), has added three AI ultrasound machines to its radiology and medical imaging service. They track breast lesions live and advise the doctor on how suspicious findings are, help identify and classify thyroid lesions, and estimate liver fat percentage and elasticity. Each costs USD 67,000, part of a renewal replacing the service's seven older scanners, which see about 950 patients a month.
Spitalul Judeţean de Urgenţă Piatra-Neamţ — Romania's Piatra-Neamţ County Emergency Hospital has placed an AI hologram at the doorway of its specialist outpatient department. The virtual assistant gives patients doctors' names and hours, specialities, consulting rooms, wards and the papers needed for a consultation, to ease routine questions at reception. A day after launch, local paper Mesagerul de Neamţ saw patients walk past it to the desk, because the hologram can talk but cannot book appointments.
Secretaria Municipal de Saúde de Feira de Santana — Feira de Santana's municipal health department launched Cuidado em dIA on 2 September, an AI clinical decision support platform from startup Volts helping the city's doctors weigh clinical data for diagnoses, treatment and referrals. The decision aid was presented to network doctors and nurses; final clinical decisions stay with professionals.
Brazilian Ministry of Health — Brazil's Ministry of Health has unveiled, at the Hospital das Clínicas of the University of São Paulo's medical school, a command centre for the SUS's intelligent ICUs, offering live AI ICU monitoring of vital signs in six hospitals in Manaus, Recife, Brasília, Belo Horizonte, Rio de Janeiro and Porto Alegre. The network has 60 active smart beds under round-the-clock watch, and the ministry aims for 280 in 14 institutions. Health minister Alexandre Padilha described it as genuine artificial intelligence working inside the public system.
Tartu University Hospital — Tartu University Hospital ran BMVision Kidney beside routine reporting by radiologists for half a year, and AI kidney cancer detection flagged malignancies in four patients whose tumours those readers had overlooked. Clinician and software together reached 99.2 percent sensitivity, with about 800 scans crossing the system monthly. Priit Salumaa, chief executive of Better Medicine, the Estonian firm behind this radiology AI, described what was built as a second pair of eyes hunting for the disease.
Peaasi.ee — Estonia's Peaasi.ee has rolled out AI mental health counselling through Peabot, a free, anonymous chatbot available around the clock beside its human chat line, built on a Finnish model. Peabot answers only from pre-checked expert content instead of freely generated text. Project manager Anna-Liisa Leppik warns ordinary AI counselling chatbot tools can mislead struggling users with unsafe answers, but says Finland's example proves one designed with mental health specialists can offer comfort without cutting out real human contact.
Chil Group — Chil Group's new E-Ward Round Platform brings AI-powered triage in healthcare to nearly 400 rural Ugandan facilities. Remote clinicians now consult city-based specialists through shared virtual rounds, as software scans vital signs to spot and rank urgent cases automatically. Linked devices and digital records drive this AI-driven patient prioritisation.
Health Insurance Review and Assessment Service — Korea's Health Insurance Review and Assessment Service (HIRA) has started AI claims review for knee osteoarthritis, applying AI-generated medical image readings. This is HIRA's first AI use within the review of claims, going beyond clinical decision support, and human reviewers keep responsibility for each final decision.
Chelsea and Westminster Hospital NHS FT — Chelsea and Westminster Hospital NHS Foundation Trust has adopted handheld ultrasound guided by ThinkSono's AI to exclude deep-vein thrombosis (DVT), so general clinicians can scan patients at their first visit while sonographers or radiologists review remotely. In a three-month, 120-exam pilot, the trust says 97 cases were cleared of DVT and 80% of patients went home after a single scan; for those needing more tests, average time to diagnosis fell from 19 to 9 hours. No independent evaluation exists.
Queen Victoria Hospital NHS FT — Queen Victoria Hospital NHS Foundation Trust uses Harrison.ai's AI chest X-ray tool to triage scans, pushing potentially significant findings up the reporting queue. Imaging clinical lead Dr Mohamed El-Belihy said an experienced clinician still reviews every film before reporting; the tool has run since July 2026 within a wider NHS rollout.
Cleveland Clinic Abu Dhabi — Cleveland Clinic Abu Dhabi has deployed several AI medical imaging tools. Transpara highlights suspicious mammogram areas ahead of clinician review, intelligent radiotherapy software adapts cancer treatment doses mid-care, and ARTIS Icono 3D imaging aids urgent stroke intervention.
Sheikh Khalifa Medical City & Al Ain Hospital — Sheikh Khalifa Medical City and Al Ain Hospital have brought in AI sepsis detection, with algorithms and alerts flagging patients' risk of sepsis and septic shock. Catching early warning signs buys clinical teams extra time to start treatment, the hospitals say.
Al Jalila Children's — Al Jalila Children's Speciality Hospital has introduced AI seizure detection using algorithm-assisted digital EEG equipment. The system recognises patterns linked to epileptic seizures, aiming to warn early.
Nizhny Novgorod Regional Clinical Hospital and Oncology Dispensary — Nizhny Novgorod's Semashko Regional Hospital and its regional oncology dispensary have rolled out MosMedII, an AI medical imaging platform, for X-ray, CT, mammography and fluorography scans. This AI-assisted diagnostic imaging cut turnaround from 40 minutes to 15-20 minutes, aiding earlier cancer detection, the facilities report.
Uganda Ministry of Health — AI febrile illness surveillance is under trial by Uganda's Ministry of Health with partner Medic, running Audere's HealthPulse AI Interpreter in Mpigi, Moyo, Buikwe and Lira districts. Over a silent four-month stretch, Phase I let the tool operate unseen next to Village Health Teams (VHTs); outcomes revealed strong matching between computer-generated and human calls. Phase II now lets VHTs weigh their own judgement against the machine's suggestion, checking whether visible AI-assisted illness detection boosts ground-level choices.
ASST Sette Laghi (Ospedale di Circolo, Varese) — Varese's Ospedale di Circolo (ASST Sette Laghi) uses AI stroke imaging to interpret CT, perfusion and angiography scans in minutes, passing automated analysis to neurology and neuroradiology specialists for faster treatment decisions. The hospital treats about 600 ischaemic stroke patients yearly; specialists validate every AI result, and the vendor is unnamed.
Länsi-Uudenmaan hyvinvointialue — An AI scribe from Tandem Health reaches around 7,000 social and health care staff at Länsi-Uudenmaan hyvinvointialue in Finland, phased in from September 2026 to end-2027. It drafts a note from each patient meeting for the clinician to check, amend and confirm. Care, clinical judgement and the file stay the practitioner's, whatever AI-assisted charting handles beforehand.
Clinical Emergency Hospital, Bishkek — China has donated a Digital Rehabilitation Therapist (DRT) system now operating at the Clinical Emergency Hospital in Bishkek, part of a wider push for AI rehabilitation therapy in Kyrgyzstan. The device maps 35 key body points via 3D vision and motion recognition, building personalised recovery plans and tracking patients from hospital to home. Guidance draws on over 100,000 clinical cases from Peking University Third Hospital.
Calderdale and Huddersfield NHS FT — Calderdale and Huddersfield NHS Foundation Trust has built an AI clinical decision-support tool, the Anticholinergic Medication Index, into its electronic patient record for safer prescribing in patients 65 and over. From 1 September 2026 it automatically recalculates an anticholinergic-burden score from 88 commonly prescribed medicines, flagging risks such as falls and delirium.
Te Whatu Ora (Health NZ) — Health New Zealand (Te Whatu Ora) is rolling out Heidi, an AI medical scribe turning consultations into clinical notes, to about 1,000 emergency-department clinicians nationwide after trials including Hawke's Bay Hospital. Hawke's Bay trial clinicians reported saving about 11 minutes per consultation on average.
12 health systems — Twelve US health systems, among them Advocate Health, Cedars-Sinai, Houston Methodist, Mount Sinai, Northwell Health and Sutter Health, are jointly adopting Aidoc's CARE foundation model and aiOS platform for urgent-imaging triage. They serve nearly 20 million patients yearly and aim for faster reads and sooner critical results, with outcomes due in 2027.
LCMC Health — LCMC Health is deploying Qualified Health's HIPAA-compliant generative-AI platform across operations, revenue cycle and care delivery enterprise-wide, for about 15,000 clinicians and staff. The platform offers performance monitoring, access controls, patient-consent safeguards and automated agent building, moving the system from limited pilots to full use.
MGM China — MGM Cotai has installed a flagship TCM AI health station giving staff on-site AI-assisted Traditional Chinese Medicine screening. Devices read biometric signals from the tongue, face and pulse to flag over 500 possible conditions. The Macau Federation of Trade Unions, the Macao Society for Music Promotes Health Research and Kiang Wu Hospital jointly run the scheme for workers in Macao.
OSF HealthCare — OSF HealthCare extended RapidAI's Rapid Enterprise Platform to all 18 of its hospitals; its AI reads imaging and flags suspected strokes, and supports oncology, trauma, vascular, cardiac and neuro care. OSF says its first rollout measurably sped treatment and made stroke identification more consistent; no independent evaluation exists.
Novo Nordisk — Novo Nordisk and Amazon Web Services have formed a strategic partnership for agentic AI in drug discovery, opening a co-innovation hub at Novo Nordisk's King's Cross site in London to speed the journey from target to first-in-human dosing. AWS tools including Amazon Bedrock and Amazon Bio Discovery will help find targets and design therapies.
Uganda Ministry of Health — Uganda's Ministry of Health has launched Nurse Lina, a WhatsApp AI chatbot for family planning giving immediate, private answers on contraceptive choices, side effects and closest service points. This family planning assistant uses ministry guidelines and common queries within a wider reproductive health information push. It improves with use, referring users to face-to-face care where needed.
MediMind — MediMind now runs an AI-native hospital information system across 25 hospitals in the Republic of Georgia, part of a growing wave of AI in hospital management built on the FHIR R4 data standard with Medplum. Workflows span admission to discharge, with roughly 1.1 million patient records, 3.1 million visits and 3.7 million clinical observations already logged. Algorithms flag preliminary findings and decision-support alerts, but this machine-assisted clinical documentation stays labelled as AI-generated, never folded into a clinician's final conclusions.
Kingston & Richmond NHS FT — Kingston & Richmond NHS Foundation Trust is rolling out Lyrebird Health's ambient voice technology, turning clinical conversations into notes, from early August 2026, starting in the emergency department and gastroenterology and gynaecology outpatients. After a ten-specialty pilot, the trust wants the AI note-taking to lighten admin and free time for care.
Brazilian Ministry of Health — Brazil's health ministry runs its National Network of Intelligent ICUs, AI in intensive care backing clinicians, in six hospitals: Manaus, Curitiba, Porto Alegre, Recife, Belo Horizonte and Rio de Janeiro's Hospital Universitário Clementino Fraga Filho. Companion pilot SAMU Inteligente links live clinical data across emergency dispatch centres, ambulances and intensive care units in Recife, Belo Horizonte and Porto Alegre. It set out the smart ICU network at an intelligent-hospitals symposium.
National University Hospital — Singapore's National University Hospital has put Ainex's ENAD real-time AI endoscopy system in all its endoscopy units to find and characterise gastrointestinal lesions in upper and lower GI procedures. The rollout follows HSA approval; Ainex says a 998-patient multicentre randomised trial found ENAD-assisted colonoscopy lifted the adenoma detection rate from 36.1% to 52.3%.
Unity Health Toronto — Unity Health Toronto's AI-assisted randomised trial with the GEMINI research network aims to predict and prevent delirium in 13 Ontario hospitals, funded by Ontario's Ministry of Health. The trial runs to March 2027 with about 15,000 patients and 2,000 frontline staff. An AI tool flags the highest-risk patients for prevention; researchers say it identifies delirium at about triple historical accuracy.
Rakuwakai Marutamachi Hospital — Rakuwakai Marutamachi Hospital in Kyoto, Japan, is co-running a pilot with startup peaceful of NURVIS, a wearable AI agent for AI nursing documentation and bedside medication guidance. Paired with bone-conduction earphones and a smartphone, NURVIS drafts automated nursing notes. The aim: lighter paperwork and veteran know-how handed to juniors.
Macao Health Bureau (SSM) — Annual capacity tripled at Macao's Health Bureau (SSM) once AI-assisted diabetic retinopathy screening began under AI healthcare imaging, slashing per-patient exams from 15 minutes to 2, Alvis Lo said. An AI medical image reading system runs alongside a piloted AI-assisted diagnosis system, an AI diagnostic tool aiding physicians' clinical calls. Officials will fold these into the Macao One Account health platform.
Prefeitura de Porto Alegre (Secretaria Municipal de Saúde and Procempa) — Porto Alegre's municipal IT company Procempa has built MedGenIA, a generative-AI health app, into Gerint, the city health department's SUS hospital admissions system, for AI hospital admission triage. From each facility's request data it drafts a structured case summary and a suggested priority rating for the regulating physicians allocating beds.
Egypt Healthcare Authority — Egypt Healthcare Authority plans Africa's first virtual hospital in Ismailia, 119 digital units strong, for AI triage in healthcare nationwide. It splits into 32 electronic triage stations, 27 telecardiology setups, 12 teleradiology posts, remote ICU beds, virtual specialty consultations and live chronic-patient tracking, feeding into Egypt's universal health insurance scheme. Ahmed El-Sobky pitched it as one grid joining AI-assisted medical triage nationwide.
Agel — Agel has installed certified AI colonoscopy support at 12 gastroenterology centres across Slovakia and the Czech Republic. MAIA Labs software marks suspicious lesions on live endoscope video. Agel-cited clinical studies credit AI-assisted colonoscopy with a near-40% Adenoma Detection Rate gain and 61% more precancerous lesions found per examination.
Ministry of Health of Rwanda — An AI-enabled camera now links doctors at Nemba Hospital in Gakenke District with specialists based at Rwanda's Ministry of Health headquarters, forming the backbone of a new AI telemedicine for maternal health initiative meant to cut deaths during childbirth. Rather than requiring travel, this AI-powered remote consultation in childbirth lets a distant specialist guide a local doctor through delivery complications in real time. Ten additional hospitals with high maternal and infant mortality are next in line for the rollout.
Ministry of Health (Tunisia) — Tunisia's health ministry will put an AI chest X-ray reading model, built by Ariana and Monastir university hospital teams with local technology partners, into more than 300 primary health centres, notably inland, minister Mustapha Ferjani says. Trained on regional conditions such as tuberculosis, the lung-screening AI is meant to help doctors spot lung anomalies and tumours early, with clinicians keeping the medical decision.
Ghana Ports and Harbours Authority — The Ghana Ports and Harbours Authority has commissioned a Philips Incisive 128-Slice CT Scanner with AI CT scanning at Tema's International Maritime Hospital, restoring CT scan services there. Its AI-supported diagnostics cover advanced cardiac imaging, stroke assessment and lung nodule detection.
Hospital Biprovincial Quillota Petorca — Hospital Biprovincial Quillota Petorca began SITGEQ, AI surgical management with real-time dashboards, in pre-surgery and, from 27 July, its operating theatres. GesNova Salud's system gathers clinical records, waiting-list status, theatre timetables and patient tests so teams follow each case from surgical request to operation. Hospital director Eduardo Lara Hernández highlights its data analytics and AI; the goals are better theatre use, fewer cancellations and shorter public-network surgical waiting lists. Two of the Viña del Mar Quillota Petorca health service's three high-complexity hospitals now use it.
Centro Médico de Puerto Rico (ASEM) — Centro Médico de Puerto Rico's emergency room, run by the island's Medical Services Administration (ASEM), has brought in AI-assisted CT scanning, part of roughly US$1.8 million in investment. ASEM's director says the RapidAI platform will process images and aid clinical analysis, backing radiologists and doctors while leaving clinical judgement with them.
Ministry of Health of St. Kitts and Nevis — Joseph N. France General Hospital now runs AI clinical decision support under St. Kitts and Nevis' National Digital Health Information System (NDHIS), reaching over 51,000 residents. Built locally by the St. Kitts and Nevis Robotics Association, the platform ties e-prescriptions, lab data, imaging and billing to population-level analytics and an advisory-only clinical decision support layer, per the Ministry of Health. Clinicians retain final authority.
Bishkek Emergency Medical Center — Kyrgyzstan's Bishkek Emergency Medical Center launched AI ambulance call support whose intelligent assistant coaches patients before paramedics arrive. Callers get an SMS carrying a geolocation link, plus a personal account advising from clinical recommendations, and live ambulance tracking. AI-assisted emergency calling eases dispatchers and locates patients unable to describe their position.
Ghana Health Service — PROMPTS, an AI maternal health platform nicknamed Nurse Mary from Ghana Health Service and Jacaranda Health, now serves over 7,000 mothers at Greater Accra Regional Hospital. Via SMS, pregnant women and new mothers submit queries sorted by urgency, answered or forwarded to a clinic. It extends AI-driven maternal care tools across Ghana.
Wezi Medical Centre — Wezi Medical Centre has Malawi's first wireless electrocardiogram (ECG) system with AI heart attack detection, a gift from Taiwan's Pingtung Christian Hospital backed by Taiwanese firms. Readings reach doctors' phones or tablets immediately, where embedded software grades each trace and shows colour-coded warnings of elevated risk. The AI cardiac screening targets cardiovascular disease, a major national killer.
M-PESA Foundation and Kilifi County Referral Hospital — M-PESA Foundation and Proximie, a UK health technology firm, opened a Digital Operating Room at Kilifi County Referral Hospital for AI in surgery, targeting obstetric and neonatal emergencies in Kenya. Cloud AI and augmented reality let outside experts direct local teams live, cutting transfers. The Uzazi Salama Programme runs this AI-powered digital operating room for a year to lower referrals and widen specialist access on the coast.
Governo do Paraná (Hospital do Câncer de Londrina and Hospital São Vicente, Guarapuava) — Paraná has become the first Brazilian state to place AI in cancer care at SUS hospitals, helping doctors work through complex cases. First-stage sites for Projeto Capricórnio, built by Google Cloud with the Princess Máxima Center, are the Londrina Cancer Hospital and Guarapuava's Hospital São Vicente, where an oncologist says AI-assisted literature reviews take 15 to 20 minutes. Decisions remain with medical teams.
Kyushu Cancer Center — Ubie's generative AI hospital software is now live at Kyushu Cancer Center, marking the first such rollout among National Hospital Organisation (NHO) facilities. Charting for new patients takes 10 fewer minutes each, and effort spent writing up minutes has dropped 70%. Engineers at the hospital devised their own linkage mechanism joining the tool to electronic medical records, so nothing leaves the protected network — an approach to AI in medical record-keeping built entirely in-house.
Aizawa Hospital — Trials are over at Aizawa Hospital: Ubie's generative AI platform now runs facility-wide, a medical AI deployment spanning more than 30 distinct workflows in Nagano Prefecture. An added iPhone app lets clinicians dictate notes and generate paperwork automatically. Meeting records that previously demanded 60 minutes to produce are finished in 10 — a sixfold gain from this hospital generative AI platform.
Benin Ministry of Health — Benin's Ministry of Health has switched on AI malaria vector control in six communes: Copargo, Djougou, Tchaourou, Ouidah, Abomey-Calavi and Cotonou, about 336,000 residents. Geospatial mapping plus artificial intelligence flag probable mosquito breeding pools; drone crews douse the water with larvicide. Health Minister Benjamin Hounkpatin ties drone larviciding to tighter aim and quicker turnaround. Japan bankrolls the 2.3 million dollar pilot with UNICEF and the World Health Organization.
Ministry of Health, Sierra Leone — Sierra Leone's Ministry of Health, working with the Gates Foundation and China CDC, has launched an AI mosquito surveillance device that snaps images, sorts mosquitoes by species and tallies numbers on its own, instantly. The three-year initiative's first stage kicks off 1 July 2026, arming officials with a real-time activity map to sharpen where vector control efforts land. This automated insect-borne disease monitoring approach could reshape how the country tracks malaria risk.
Centro Oncológico Nacional Dr. Juan Ignacio Gutiérrez Sacasa (Ministerio de Salud, Nicaragua) — Nicaragua's Government will bring AI-assisted cancer diagnosis to Managua's National Oncology Centre once its second building phase completes on 31 July 2026. New imaging and pathology machines will carry built-in artificial intelligence, sharpening readings of X-rays and tissue samples. The move forms part of a wider push to strengthen public healthcare through AI diagnostic tools in oncology.
Sickle Cell Cares Foundation Inc. — Sickle Cell Cares Foundation Inc. (SCCF) put SickleConnect live on World Sickle Cell Day, billing the free app at app.sicklecellcares.org as the Caribbean's first digital health platform for Sickle Cell Disease patients. Logged pain, triggers, medicines, hydration and lab results feed an AI clinical summary a patient hands to a doctor. That AI health report spans 30 days, is built from patient-entered data, goes to a haematologist, and costs nothing for the first 30 days. Sickle Cell Disease reaches an estimated 1 in 500 people of African or Caribbean descent, yet specialist haematology care stays scarce in Dominica and neighbouring Eastern Caribbean states; SCCF's founder and chief executive, Kellyn George, has HbSS Sickle Cell Disease herself.
Swiss Medical Network — Swiss Medical Network, a leading private hospital group in Switzerland, is extending AI clinical documentation to every hospital, saving each doctor up to two hours a day. The Heidi platform transcribes consultations live and prepares drafts. Each AI-drafted clinical note is checked and signed off by a doctor before filing.
Côte d'Ivoire National Tuberculosis Programme (PNLT) — Côte d'Ivoire's National Tuberculosis Programme and research partner PAC-CI began trialling eHealth4Child-TB in Sassandra on 16 June 2026, bringing AI chest X-ray reading to childhood tuberculosis care. Health workers get AI-analysed radiography plus treatment decision-support algorithms, now spreading across the Abengourou, Agboville and Sassandra health districts.
National Health Insurance Company (CNAM) — Moldova's National Health Insurance Company (CNAM) has rolled out Ana, an AI healthcare assistant built into its web portal, to field citizen questions on mandatory health insurance. This AI-driven virtual assistant handles three tasks: coverage checks, payment steps, and eligible medical services. CNAM treats it as a trial run, watching results before wider use.
CHU de Poitiers — The CHU de Poitiers has made MedGPT official across all five sites, extending the French medical AI assistant to health professionals throughout the surrounding area. Billed by the hospital as a secure French solution, it answers clinical questions from reference sources.
Department of Health and Social Care — England's NHS trusts will all gain AI chest X-ray tools by 2029 under a £20 million UK government package. Half already use this lung cancer diagnosis software, which has helped over 4 million patients get faster results or all-clear news, cutting scan turnaround from eight days to four.
Pantai Hospital Kuala Lumpur — Malaysia's Pantai Hospital Kuala Lumpur, part of IHH Healthcare, introduced Varian Ethos with HyperSight for AI adaptive radiotherapy, letting clinicians reshape cancer treatment plans daily as a patient's anatomy shifts. Imaging spots anatomical changes in real time, so AI-guided radiation therapy stays on tumours and spares healthy tissue.
Ministry of Health of Bhutan — Bhutan's Ministry of Health will place Cardiolife's AI ECG interpretation software in 25 health facilities - national, regional, district - after Health Minister Lyonpo Tandin Wangchuk backed the plan. Software and kit arrive free; clinicians read ECGs instantly. Cardiolife rates its CE-marked MDR Class IIb product at up to 100% accuracy on key cardiac conditions, a vendor figure with no independent evaluation published. Validation comes through a joint operational research study of the AI-assisted cardiac diagnostics.
Tulsipur Sub-Metropolitan City — Tulsipur Sub-Metropolitan City in Nepal's Dang district is using AI TB screening in 17 of its 19 wards, part of a drive to make the city TB-free. Health officials say the AI X-ray machine spots tuberculosis infections that regular microscopy misses.
Birat Medical College Teaching Hospital — Birat Medical College Teaching Hospital in Morang runs an AI-enabled GE Signature ultrasound machine, by its own claim Nepal's first. AI ultrasound software spots organs and handles measurements, trimming scan duration by up to 40% and adding 3D and 4D views. Staff expect AI-assisted scanning to spare easterners journeys to Kathmandu or overseas for complex tests.
Ministry of Health and Social Protection — Morocco's health ministry is piloting AI ultrasound to help staff perform and read scans, mainly for pregnant women in rural areas, health minister Amine Tehraoui told the House of Representatives. The AI-assisted scanning runs at the Aghbala health centre in Béni Mellal province and Moulay Bouazza in Khénifra province, backing midwives and other health professionals, sharpening diagnosis and catching high-risk pregnancies sooner.
Ministry of Health and Welfare — Korea's Ministry of Health and Welfare will introduce AI patient referral and return in hospitals during the latter half of the year, beginning with pilots in Seoul and Gyeonggi, Gangwon and South Jeolla. Linking AI directly to each hospital's picture archiving and electronic medical record systems will automate sending patients onward and back where care happens. The broader Healthcare AI Transformation Sprint programme begins in June.
Pohjois-Pohjanmaan hyvinvointialue (Pohde) — Pohde, the North Ostrobothnia wellbeing services county, extended its AI phone assistant for cancelling or moving adult dental appointments to eight more municipalities on 26 May 2026, after Ii, Pudasjärvi, Vaala, Utajärvi and Muhos. It calls itself the first Finnish wellbeing region with an AI-based phone service in dental care, freeing professionals for other work, within the Sitra-funded UusiAika project.
Hospital Clínico Universitario (UCV) — UCV has relaunched AI in robotic surgery in Venezuela, pairing a multi-arm robot with AI tools and a digital platform linking records to booking. Acting President Delcy Rodríguez opened this AI-assisted surgical setup, spanning urology, gynaecology, thoracic and general procedures with Chinese-made equipment, for wider rollout across state hospitals.
Interact Club of Brick by Brick — Suriname's Interact Club of Brick by Brick has built Flowvie, a free AI menstrual health assistant offering young women online guidance on periods and wellbeing. Extending the club's Bloedserieus initiative, it draws answers from screened material — input from international health bodies and checked educational content — as an AI-driven menstruation guidance tool designed for reliability.
Lafto Specialized Hospital — East Africa's first AI-powered hospital, Lafto Specialized Hospital, opened with 500 beds in Addis Ababa, Ethiopia. Oncology and heart patients get an AI-assisted PET unit and 3D mammography, while lab testing runs fully robotic — medicine built on AI in healthcare.
SEHA Corniche Fertility Centre (PureHealth) — SEHA's Corniche Fertility Centre in Abu Dhabi, part of PureHealth, is using Vitrolife's iDAScore for AI embryo selection through the EmbryoScope time-lapse system, a first in the PureHealth network. Learning from embryos with known outcomes, it ranks implantation potential across full time-lapse development, replacing embryologists' manual morphokinetic analysis and putting the most viable first in line, SEHA says.
Ophthalmology Specialist Hospital (Yangon) — Ophthalmology Specialist Hospital (Yangon) ran AI vision screening during field eye check-ups in Mayangone Township, part of a Ministry of Health drive across Yangon Region to curb rainy-season illnesses. Myanmar's Union Minister of Health attended the campaign event, where the hospital director and ophthalmologists used AI-assisted eye examinations for on-site diagnostic testing. The initiative forms one strand of a wider seasonal disease-prevention effort covering all Yangon townships.
Heilbrigðisstofnun Suðurlands (HSU) — Heilbrigðisstofnun Suðurlands (HSU) has fitted Móberg nursing home with TETON AI fall detection sensors that monitor movement and sleep patterns, flagging hazards such as a resident sitting up in bed before a fall happens. Staff response times have fallen by up to 95%, per HSU. Cited research shows falls dropping 40% on average, with incidents down as much as 75% since the elderly care monitoring system began operating in Iceland.
Zanzibar Ministry of Health (ZAMEP) — Zanzibar's Ministry of Health, through the Zanzibar Malaria Elimination Programme, has launched AI drone mosquito control to find and treat breeding sites, with WHO, Japan's SORA Technology and the Ifakara Health Institute, funded by the Government of Japan. Drone surveillance reaches places conventional methods miss and lets health teams treat breeding sites on detection, officials say. Malaria prevalence on the islands is 0.04 percent.
Shuklaphanta Municipality — Nepal's Shuklaphanta Municipality, in Kanchanpur district, flagged 42 suspected TB cases among 688 people checked through AI tuberculosis screening with portable X-ray equipment. Camps in high-risk wards referred those cases for further examination. The AI X-ray analysis reads each image digitally to spot possible TB infection.
Ministry of Health of South Sudan — South Sudan's Ministry of Health launched the WHO-backed Digital Care Companion (DCC), giving frontline health workers in remote Maiwut County, Upper Nile State, AI clinical decision support under the Minister's Special Initiative on Maternal and Newborn Health. The DCC, a symbolic AI diagnostic aid using ICD-11 standards, works fully offline in low-literacy settings, guiding diagnosis and management.
Bach Mai Hospital — Hanoi's Bach Mai Hospital began free AI lung cancer screening on 10 May 2026, with AI helping read the images. The hospital calls it a 'divine eye' that detects small lung nodules and scores their risk on the international Lung-RADS scale. On day one it flagged two of 75 attendees with suspected malignant lung lesions.
Clalit Health Services — Clalit Health Services' new in-house AI PRO applies AI risk prediction to millions of medical records, flagging at-risk patients sometimes before any doctor visit and issuing about 200,000 personalised treatment recommendations monthly. A predictive-screening extension shown to over 800 clinicians at a Tel Aviv conference aims at spotting chronic kidney disease and bowel cancer early from background conditions and test-result changes. Clalit keeps final decisions with the treating doctor.
Comunidad de Madrid — Spain's Comunidad de Madrid has installed AI cancer diagnosis tools in several public hospitals to spot suspected prostate and breast cancer faster and more accurately, and to follow detected cases. For prostate cancer they help radiologists read MRI and review a patient's whole imaging history together; for breast cancer, systems reading 2D mammograms and 3D tomosynthesis flag suspicious lesions and score how likely malignancy is. Clinicians will test them on diagnosed cases in coming months, with phased installation from summer.
Turks and Caicos Islands Government — Turks and Caicos Islands Government will introduce AI case management overseas treatment for its Treatment Abroad Programme (TAP), built by the Ministry of Innovation, Technology and Energy after the plan won a spot in the World Bank Digital LAC AI Accelerator. Minister E. Jay Saunders said the tool will speed workflows, cut delays and strengthen accountability for TAP's overseas medical cases, a programme already above 10% of the national budget. This marks agentic AI in healthcare administration for a programme that consumes over a tenth of public spending.
Uganda Ministry of Health — Uganda's Ministry of Health has introduced AI handheld ultrasound in Kampala, so midwives and nurses can examine pregnant women and read results without advanced radiology training. Nicosam Healthcare delivered fifteen Gates Foundation-funded devices; over 300 health workers are trained, with at least 600 targeted by the end of 2026. Midwives at lower-level facilities can consult referral-hospital specialists on AI-assisted scans; the ministry says only about 28% of pregnant women in the country currently get even one ultrasound.
Cleopatra Hospitals Group — Cleopatra Hospitals Group cut first-treatment wait times from 38 to 23 days across 11 Cairo and Alexandria sites using an AI cancer diagnosis platform from Paige.ai, Microsoft Egypt and Sehha. It screens referrals in Arabic and English, reads pathology images by machine, and streamlines multi-disciplinary scheduling. The American University in Cairo Faculty of Medicine verified the gains across 92,000 episodes, an early case of AI-assisted pathology screening in Egypt.
Helse Sør-Øst RHF — Helse Sør-Øst, south-east Norway's health authority, decided on 9 April 2026 on a framework agreement with four AI note-taking suppliers and a binding order for 4,000 licences. Their speech-to-summary tools transcribe clinician–patient talk and draft a structured record summary. The authority will brief its board on wider rollout in autumn 2026, after early hospital use.
Jakaya Kikwete Cardiac Institute — The Jakaya Kikwete Cardiac Institute in Dar es Salaam opened a laboratory with AI lab automation and robotic sample processing, cutting test-result waits from three hours to one, chief executive Peter Kisenge says. Samples are analysed precisely, and the lab doubles as a live training site for local health professionals.
Botswana Medical Aid Society — Botswana Medical Aid Society's 90,000 members can now unlock a face-scan health score via Thrive with Bomaid, an AI wellness app on ordinary smartphone cameras. Over 70 biometric inputs, spanning heart rate, stress and blood pressure, generate a VivaScore, paired with automated nudges, reports and goals inside this AI health scoring platform. Debswana and Engen tested it for three months first.
Government of Papua New Guinea (National Department of Health) with World Vision — Papua New Guinea now runs AI tuberculosis screening through four battery-powered X-ray units deployed by the government with World Vision under the Systematic Screening Initiative. Health workers carry backpack-sized devices built on CAD4TB software into areas without hospital infrastructure, where each scan takes 10 seconds to produce a score flagging presumptive cases, including in people showing no symptoms. One unit operates as a mobile setup; three are fixed at Gerehu General Hospital, Six-Mile Clinic and Lawes Road in Konedobu, giving the country's AI-assisted TB detection reach beyond urban centres.
Chad Ministry of Public Health and Prevention — Health authorities in N'Djamena are piloting Vacci360, an AI vaccination system that extends immunisation to remote corners of Chad. Built atop DHIS2, it flags where terrain limits clinic reach, auto-generates mobile team routes, tracks shifting human and livestock populations, and guides staff through dense data via chat. Lac, Kanem and Hadjer Lamis provinces run the scheme, funded by the Gates Foundation, WHO and UNICEF.
Secretaría Distrital de Salud de Bogotá — Bogotá's health department runs Proyecto STOP, AI suicide risk detection that scans social media for patterns tied to depression, suicidal ideation and other mental-health risks. The system links people at risk to Línea 106, the department's free emotional-support line, for earlier prevention.
Belarus Ministry of Health — Belarus's Ministry of Health will put AI chest X-ray reading into Minsk healthcare institutions from the middle of 2026, for digital chest radiography. Elena Krotkova, who directs the Republican Scientific and Practical Centre for Pulmonology and Phthisiology, says trained neural networks will stop specialists overlooking minimal pathological changes on an image. A radiologist still reads and signs off every picture, and the ministry intends AI radiology support to reach all regions later.
eHealth Africa and Jacaranda Health — eHealth Africa and Jacaranda Health extended PROMPTS, an SMS-based AI maternal health platform, to Kano State for 25,000 mothers in underserved communities. Hausa and English texts deliver pregnancy and newborn guidance to phones. Nigeria follows Kenya, where the same AI-enabled maternal health messaging tool had already reached over 3.8 million mothers.
Ministerio de Salud de Río Negro — Río Negro's provincial health ministry will pilot SAP-RN, a Google Gemini tool for AI emergency triage, for 90 days in Viedma and Cipolletti hospitals. The priority-care system will also track waiting times, resolved cases and 72-hour re-attendance to plan emergency staffing, with clinical judgement keeping the final word.
IMSS-Bienestar — IMSS-Bienestar has opened the Hospital Oncológico para la Mujer in Mexico City, with AI mammography reading the images. Director general Alejandro Svarch Pérez set out fixed timings: appointment and linking of records to the AI mammograph on day zero, medical assessment, mammogram, ultrasound and further studies on day one, results by day 14 and treatment from day 30. A robotised pathology laboratory and an AI innovation centre offering assisted image processing complete the site.
National Hospital of Sri Lanka — The National Hospital of Sri Lanka has added an AI MRI machine to patient care, the first in a government hospital nationwide, the Health Ministry says. Worth more than Rs. 450 million, the AI-equipped scanner serves the Emergency Treatment Unit and Neurosurgery Unit and was introduced under the patronage of Health Minister Nalinda Jayatissa.
Armauer Hansen Research Institute / Ministry of Health — AI-PRESCRIBER decision support in Ethiopian hospitals begins operating at 13 sites after specialists finished joint instruction from the Armauer Hansen Research Institute and Ministry of Health. Each facility sent one trained clinician back to embed the AI clinical decision support system into routine care, aimed at lifting adherence to Standard Treatment Guidelines.
Kenya Ministry of Health — Kenya's Ministry of Health will put AI in community health through Causal Foundry, an adaptive-intelligence specialist in medical care. Causal Foundry is to draw patterns from live household records gathered by the 100,000 community health promoters given smartphones by the state, directing care and resources where demand is highest. Ministry officials have already discussed rollout with the firm.
UZ Gent — Ghent University Hospital has appointed the first full-time AI manager at a Flemish hospital to run at least 30 hospital AI projects, live or planned. In live use, AI locates the clot on CT scans in suspected stroke, replaces one of the two doctors reading each breast-screening scan, and does pre-surgical tissue-structure marking in minutes rather than hours. A doctor always has the final say, and the hospital discloses whenever an AI tool was used.
Uganda Cancer Institute — Uganda's national programme against cervical cancer, operated by the Uganda Cancer Institute and Ministry of Health alongside the Korea Foundation for International Healthcare, will roll out AI cervical cancer screening with AI-assisted visual inspection (AI-VIA) devices, expected to speed and sharpen detection. Its three-year pilot has screened over 35,000 women aged 25 to 65 in Kampala and Mbarara, treating more than 3,000 with pre-cancerous lesions.
Semmelweis University and the Belügyminisztérium — The Belügyminisztérium in Hungary has opened a national AI dementia screening study called 'Mesterséges intelligencia alapú digitális kézmozgás-analízis alkalmazása a kognitív hanyatlás korai felismerésében', guided by Semmelweis University with the Országos Kórházi Főigazgatóság. Kamondi Anita's team at the Department of Neurosurgery and Neurointervention runs it across ten health institutions. Validating the tablet-based cognitive decline test on 3,000 people precedes any move into surgeries.
CHU de Martinique — CHU de Martinique now uses AI stroke thrombectomy support via RAPID ANGIO, its tool for AI-based cerebral scan analysis, in the interventional neuroradiology suite. During mechanical thrombectomy for ischaemic stroke, it reads scans fast, marks damaged zones and flags tissue worth saving. Roughly 900 stroke patients need hospital care in Martinique yearly.
Chad National Tuberculosis Control Program (PNT) — Chad's National Tuberculosis Control Program (PNT) used AI-powered X-ray screening to find 112 TB cases among nomads, prisoners and refugees. The CAD4TB tool, mounted on a mobile truck, examined 7,116 people across 12 detention sites and 9 refugee camps, flagging 589 abnormal scans. This AI tuberculosis screening effort also trained 220 health staff, 90 volunteers and 120 prison cell leaders.
Thika Level 5 Hospital / Kiambu County Health Department — Thika Level 5 Hospital is testing MedRef AI clinical decision support, guiding Kiambu County clinicians through Ministry of Health protocols via instant lookup. Staff train on it alongside normal duties. Wider rollout to other Level 4 and Level 5 facilities hinges on how well this AI-powered medical reference system performs.
Fiji Ministry of Health and Medical Services — Fiji's Ministry of Health and Medical Services is fitting public hospitals with AI-powered X-ray machines, swapping out analogue equipment over 15 years old. The digital upgrade uses AI in medical imaging to help clinicians catch abnormalities and reach diagnoses faster. Colonial War Memorial Hospital in Suva, Nausori Health Centre and Taveuni are already equipped, with 21 sites planned nationwide.
Grenada General Hospital — Grenada General Hospital provides AI breast cancer screening via a Hologic 3D digital mammography system, the public healthcare system's first mammography service. Donated by US radiologist Dr Randy Becker and installed in January 2026, the unit combines 3D tomosynthesis with artificial intelligence detection tools. This AI-assisted mammography lets radiologists catch cancers sooner.
BPJS Kesehatan — BPJS Kesehatan, which runs Indonesia's national health insurance scheme, has launched four AI services, including SmartClaim JKN, AI claims verification that speeds its verifiers' work. SISCA JKN, a digital avatar, gives Mobile JKN app users information and first responses, while Smart Analytics and Smart Insight turn agency data into material for operational and managerial decisions. IT director Edwin Aristiawan presented the launch.
Hôpital Foch — Hôpital Foch in Suresnes, near Paris, has deployed Tandem Health's AI medical scribe, which drafts consultation reports automatically from doctor–patient conversations. Used in close to a thousand consultations already, it went first to documentation-heavy specialties such as neurology and urology, the hospital says.
Social Health Authority — Kenya's Social Health Authority vets health insurance claims on arrival with AI fraud detection that highlights irregularities, Health Cabinet Secretary Aden Duale informed the Senate Health Committee. The fraud engine exposed a Kwale County member enrolled with 381 dependent children and monitors service trends at each facility, spotting patients logged for dialysis ten times a day. Suspect cases go to the Directorate of Criminal Investigations.
Tachibana Hospital — Tachibana Hospital, an orthopaedic specialty centre in Japan's Miyazaki Prefecture, is running the country's first orthopaedic-hospital pilot of OPTiM's generative AI platform, OPTiM AI Hospital, applying AI-assisted clinical documentation to cut paperwork time. The platform draws on electronic medical records to produce nursing summaries, clinical forms and physician opinions, with a targeted 30 to 50% reduction in drafting time. This automated hospital record generation marks an early test of how specialty hospitals might offload routine record-keeping.
Nicaragua Ministry of Health — Central America's first public genetics facility opened on 28 January 2026 when Nicaragua's Ministry of Health inaugurated the National Genetics Centre "Blanca Segovia Sandino Aráuz" in Managua. Its opening phase runs on AI microscope genetic diagnosis, pairing a high-resolution microscope with embedded artificial intelligence to speed up karyotyping and spot chromosomal disorders and leukaemia locally, cases that once meant travelling abroad for care. Officials plan to broaden services later toward molecular biology and biochemical genetics, extending beyond this initial AI-assisted genomic analysis capability.
Ministry of Health Brunei Darussalam — Brunei Darussalam's Ministry of Health, the Novartis Foundation and EVYD Technology used Davos to pledge AI cardiovascular risk tiering for a whole country. Leaning on the state's own medical records, this population-wide heart disease prevention drive promises earlier detection of the most endangered residents and a fitting remedy. No country has taken CARDIO4Cities national before, and Brunei will cover four hazards at once — hypertension, diabetes, obesity, hyperlipidemia.
Selangor state government — Selangor's state government will pilot 150 AI fall detection devices in elderly care centres and private homes from June 2026, through state foundation Yayasan Warisan Anak Selangor with South Korean tech firm UNIUNI. Infrared sensors track movement, and if someone falls the system detects it fast and alerts care-centre operators or others responsible for immediate help, says foundation chief executive Gan Pei Nei. Devices go in only with residents' consent.
Univerzitná nemocnica Bratislava (UNB) — Central Europe's first AI-assisted brain surgery took place at Univerzitná nemocnica Bratislava (UNB), where neurosurgeons used an AI model from the University Hospital Rio Hortega in Valladolid, Spain to scan intraoperative ultrasound in real time, flagging tumour remnants invisible to the eye. Post-operative MRI showed no residual tumour, and the patient was discharged. The case marks an early regional milestone for AI brain tumour surgery.
Instituto Nacional de Enfermedades Respiratorias (INER) — Mexico's National Institute of Respiratory Diseases (INER) has brought in AI imaging software to catch lung cancer and conditions such as pneumonia, COPD and tuberculosis earlier, the health ministry says. Its head of oncology says AI chest X-ray reading gives non-radiologist physicians automated support, with CT scans next.
Ministerio de Salud del Perú (Minsa) — Peru's health ministry will modernise radiology at the María Auxiliadora and Cayetano Heredia hospitals and Lima Norte network facilities with AI radiology diagnostic support for more than four million people. Health minister Luis Quiroz accepted the equipment from AstraZeneca Peru under a cooperation agreement, for faster remote breast diagnosis within the ministry's Digital Hospital.
Ministerio de Salud de Panamá (Minsa) — Panama's Ministry of Health is piloting VANESSA, an AI dengue surveillance platform, for six months in San Miguelito. Phase one is a WhatsApp bot where residents report suspected dengue cases and mosquito breeding sites with location, photos and text, backed by regional health-promotion, epidemiology, environmental-sanitation and vector-control teams.
MUSC Health — MUSC Health's 'Emily', a voice AI agent on SoundHound's platform linked to Epic, answers patient calls on prescription refills and appointment booking. MUSC says it has handled over 2.2 million calls, and in 2026 it expanded to retail and specialty pharmacy refill requests and status checks.
Kaiser Permanente's Permanente Medical Group uses an ambient AI scribe that drafts clinical notes from consented patient visits straight into the record. A peer-reviewed one-year NEJM Catalyst evaluation found 7,260 physicians using it across 2,576,627 encounters from October 2023 to December 2024, saving about 15,791 documentation-hours, one of the largest reported.
UpDoc is the first Software as a Medical Device the FDA has cleared with a patient-facing large-language-model interface (510(k) K253281, 23 December 2025): a prescription insulin-management tool for adults with type 2 diabetes, its LLM front-end sitting over deterministic dosing logic. It disproves the oft-repeated line that the FDA had authorised no generative-AI medical device, though reviewers call its 'AI doctor' marketing overblown for a bounded, single-purpose device.
City Clinical Hospital No. 1, Shymkent — City Clinical Hospital No. 1 in Shymkent added CEREBRA, AI stroke detection that reads brain CT scans for acute cerebrovascular events, in July 2026, continuing a phased city rollout that began at City Hospital No. 2. Its doctors treat it as a supplementary tool and still decide the stroke type themselves; over 4,000 studies went through CEREBRA across Shymkent in the first half of 2026.
Metro South Health — Metro South Health began deploying Heidi's ambient AI scribe in June 2026, a rollout running through the following year. The scribe hears consultations and writes clinicians' clinical documentation for them, after a 2025 trial in which 92 per cent of participating clinicians spent less time on administration.
National Health Security Office — Thailand's National Health Security Office board made AI chest X-ray reading a health benefit in April 2026, bringing it into real use across 610 public health network hospitals. SiAiR was built at the Faculty of Medicine Siriraj Hospital with Thai firm Perceptra and Hewlett Packard Enterprise. It relieves provincial hospitals short of specialist radiologists so remote patients get results as accurate as the capital's.
Region Hovedstaden (now Region Østdanmark) — Hospitals in the Capital Region of Denmark have used an AI fracture detection algorithm on emergency X-rays since January 2026, supporting decisions by junior doctors and treating nurses. It runs at emergency units including Bornholm, Hvidovre, Amager, Frederikssund, Helsingør, Hillerød, Frederiksberg, Bispebjerg and Glostrup, and will cover all of the future Region Østdanmark. The goal is shorter waits and quicker discharge for patients without fractures.
Valstybinis patologijos centras (Vilnius University Hospital Santaros Klinikos) — Lithuania's State Pathology Centre, part of Vilnius University Hospital Santaros Klinikos, has adopted Ibex Breast HER2, AI digital pathology that helps pathologists score the HER2 biomarker in breast cancer. Santaros Klinikos calls it the first CE-IVDR certified algorithm of its kind in the Baltic states.
Военномедицинска академия (Military Medical Academy), Sofia — The urology clinic of Bulgaria's Military Medical Academy has introduced VENUS, AI surgical navigation that segments the target organ automatically, and performed four highly specialised procedures on day one. The academy calls itself the country's first medical centre using it.
Ethiopian Ministry of Health — Ethiopia's Ministry of Health and US-based Last Mile Health have introduced HEP Assist, an AI assistant for health extension workers serving remote communities, across seven regions and 15 woredas. Answers come solely from 100 Ministry of Health guidelines through retrieval-augmented generation. Officials credit the workers' AI tool with over 6,000 consultations since January 2025, initially via call agents, now hands-on.
Caja Costarricense de Seguro Social (CCSS) — Costa Rica's social security fund, the Caja Costarricense de Seguro Social (CCSS), is using AI waiting list management to update and prune lists for surgeries, procedures and outpatient consultations. The bot compares the digital health record, civil registry and other records to find cases no longer needing care, and staff validate every case before anyone leaves a list.
Government of El Salvador (DoctorSV) — President Nayib Bukele has extended DoctorSV, El Salvador's digital health platform, to chronic patients with kidney disease, hypertension or diabetes, announcing its second stage in a national broadcast. Each patient gets round-the-clock AI health agents monitoring them, and doctors receive patients with history, questionnaires and diagnostic suggestions already processed by the platform's AI.
Instituto de Seguridad y Servicios Sociales de los Trabajadores del Estado (ISSSTE) — Mexico's state-employee health institute ISSSTE has opened four AI mammography reading centres, the 'Cuartos Azules', in hospitals in Mexico City, Jalisco, Yucatán and Coahuila. With six stations each, they aim to handle over 400,000 mammograms a year, interpreted remotely by breast-imaging radiologists, officials told the president's briefing.
Sheba Medical Center — ARC, innovation arm of Israel's Sheba Medical Center, deployed SmartER, an ambient AI clinical documentation platform, in its emergency department. Built with Sheba's AI Center and partner ScribeMD, the AI scribe assembles history, nursing intake and prior results pre-consultation, turns the conversation into a structured note and drafts a discharge summary for review.
Doctors Medical Centre — Doctors Medical Centre has begun AI MRI scanning for the public in Antigua ahead of a formal opening, driven by waits of several weeks elsewhere. The open MRI — described as the region's first, and built for patients who cannot tolerate an enclosed scanner — draws on artificial intelligence to sharpen image accuracy. Dr Joey John, medical director, said anyone may use the service, not solely people queuing through Mount St John's Medical Centre, with the backlog the priority. Qualified specialists staff the site, every machine is new rather than refurbished, and wider AI-enhanced diagnostic imaging is planned.
Horacio Oduber Hospital — Horacio Oduber Hospital has rebuilt AI prostate MRI diagnosis around scanning before sampling, so men with a high PSA in Aruba no longer go straight to a prostate biopsy. Where the scan flags one or more doubtful zones, fused MRI and ultrasound views show where to take tissue. Siemens Nederland and Global Medica reworked the radiology department's existing scanner for AI-enhanced prostate imaging, adding artificial intelligence for quality and efficiency plus replacement ultrasound units. Urologist Leong reports sharper targeting, less pain and fewer complications.
Waspito — Waspito opened AHIM, the African Health Intelligent Model, as an AI health assistant answering 24/7 on WhatsApp in English and French from any country. The Africa-focused platform channels queries, laboratory orders, prescription and record uploads, cover purchases and doctor referral through that WhatsApp health chatbot, while stressing AHIM supplants no clinician: consultation, diagnosis and treatment stay with qualified physicians on its roster. Entry costs nothing; billing falls on whichever care is taken up, from 3,000 FCFA with a general practitioner and 6,000 FCFA with a specialist doctor. Begun 2020 in Cameroon, Waspito reports its own totals: more than 1 million users, over 250,000 patients served, upwards of 2,000 licensed doctors.
WETECH — WETECH carried Alert GBV Lite, an Android app for AI gender-based violence reporting, into 11 communes of Cameroon's Far North via an outreach drive from 1 to 27 June 2026 with associations ADELPA and ALVFF. Its AI companion for survivors runs psychosocial orientation and structured reporting beside a geolocated SOS button capturing evidence mid-incident, a live service directory, anonymised secure case reporting, offline essentials, and French, English, Pidgin and Fulfuldé screens; without a smartphone or data, USSD code #080# takes a report. Project UNJP/CMR/061/EC, funded by DG ECHO and run by UNFPA Cameroon with FAO, covers it.
Elva Haigla — Elva Haigla became the first hospital anywhere to run DocAid, signing for the platform at the end of April 2026 and giving clinicians an AI medical scribe for the five to ten minutes a consultation otherwise loses to keying complaints and history into the patient file. A microphone captures the exchange, the software strips out chatter and returns complaints, history and the objective finding on screen within seconds, in Estonian, Russian or English; processing happens on the supplier's server, which never sees the name-and-identity-code file, so the reviewed and corrected wording is carried across by hand. Take-up is voluntary yet nearly every doctor asked for it, and nurses and physiotherapists have the same access — clinical documentation AI having previously run only in limited use at other medical institutions.
Shanghai Eye Disease Prevention and Treatment Center — Kong Pisei Hospital and Techo Sen Koh Thom Hospital in Cambodia are the first sites outside China to get an AI eye screening system, Vision Health Intelligent Management Centres from the Shanghai Eye Disease Prevention and Treatment Center. Big data analytics and the Internet of Things join the artificial intelligence behind this digital vision health management approach, spotting eye disorders sooner and keeping patients on tiered care, while Shanghai teams coach local staff on the kit. Patients topped 100 in month one and passed 500 by May.
Heves County Markhot Ferenc Teaching Hospital — Heves County Markhot Ferenc Teaching Hospital in Eger, Hungary, has put a scanner to work that performs AI cytology analysis on patient samples. Sharper findings and markedly quicker throughput are what the pathology department expects of the device. Shorter waits for a verdict mean earlier treatment, and the hospital frames machine-read cytology as movement towards digital pathology.
Semmelweis University — Semmelweis University now runs AI cephalometric analysis software inside the clinical setting of its dental faculty in Hungary. The tool weighs where a patient's skull, jawbones and teeth sit relative to one another, and the university puts assessment at under half a second. Structuring dental records around AI-read dental X-rays is meant to hand clinicians back time at the chair.
Heilbrigðisstofnun Norðurlands (HSN) — Careflux, an Icelandic firm, built Skríban, an AI medical note-taking tool now piloted by four doctors at HSN (Heilbrigðisstofnun Norðurlands) in Akureyri, turning spoken consultations into clinical notes in real time. Running three to four months, the trial has yielded informal estimates of up to 50% less paperwork time, a figure not independently verified. This AI clinical documentation tool already operates at Reykjavík's Kirkjusandur Health Clinic and in parts of Landspítali University Hospital.
Ziemeļkurzemes reģionālā slimnīca — Northern Kurzeme Regional Hospital (Ziemeļkurzemes reģionālā slimnīca) in Ventspils has adopted AI X-ray analysis via Gleamer AI, an AI-based medical imaging tool, catching bone fractures and chest abnormalities. Doctors across radiology, anaesthesiology and trauma units consult the software every day as routine diagnostic support rather than an experimental trial.
National Medical Research Centre for Traumatology and Orthopaedics (CITO) — Four operating theatres at CITO, Russia's National Medical Research Centre for Traumatology and Orthopaedics, now use a voice AI agent surgical safety checklists tool, letting surgeons respond verbally rather than filling in paperwork. Speech recognition and large language models drive the checklist automation during orthopaedic procedures. More than 1,300 operations have relied on this AI-assisted checklist tool since the project launched in June 2025.
Medfam Holdings (Rhino Park Private Hospital, Welwitschia Hospital) — AI patient experience agents now handle admissions and support at two Namibian hospitals run by Medfam Holdings, the country's largest private hospital network: Rhino Park Private Hospital in Windhoek and Welwitschia Hospital in Walvis Bay. Patients reach the system via WhatsApp or webchat in Afrikaans, English and German for intake data, complaints, surgical guidance, specialist lookups and quote requests. One shared platform links both sites, so a patient can move between hospitals without repeating details, an early case of AI in healthcare administration.
Koryo Medical General Hospital — Koryo Medical General Hospital, with Pyongyang Computer Technology University, built the Koryo Treatment Support System, an AI Korean medicine treatment tool that diagnoses internal medicine conditions by Korean medical methods and drafts herbal remedy and acupuncture prescriptions. It won a National Science and Technology Achievement Introduction Certificate in 2025. Newly built hospitals, including Kangdong County Hospital, now use it, spreading computer-aided traditional medicine diagnosis nationwide.
Oasis Imaging and MRI Centre — Oasis Imaging and MRI Centre now offers AI-enabled diagnostic imaging in St Andrew, Jamaica, cutting scan durations up to 60 per cent. Radiologists gain sharper diagnoses via MRI, CT, X-ray, ultrasound, mammography and prostate/breast scans, with MRI ready in 24-48 hours, through this AI-driven medical imaging service.
Saint James Hospital Group — Saint James Hospital Group's AI-enabled endoscopy scopes in Malta lift detection rates roughly 24 per cent, backed by an AI second reader for breast imaging. Its shift toward AI-assisted reporting in hospital settings adds an ambient system for AI-assisted clinical documentation, capturing live consultations into a finished report by appointment's end.
Pauls Stradiņš Clinical University Hospital — Pauls Stradiņš Clinical University Hospital in Riga, Latvia, has run over 8,000 patient CT scans through Aidoc's clinical AI stroke detection tool. Using a secure mobile app, this AI stroke triage has automatically alerted the stroke team to more than 500 suspected positive cases.
Institut National de Recherche Biomédicale — Anthropic's Claude now underpins AI outbreak reporting at DRC's Institut National de Recherche Biomédicale, cutting Ebola sitrep turnaround in the Bundibugyo outbreak to under an hour, down from a full day. A skill built by WHO AFRO pulls case counts from health zone PowerPoint slides, flagging emerging trends in this AI disease surveillance push. Claude Science tools also help INRB assemble viral genomes and chart family trees from sequencing data.
University Hospital North Norway (UNN) — University Hospital North Norway (UNN) in Tromsø uses AI patient monitoring on the gastro-surgical ward, where wireless sensors track pulse, respiration, blood pressure and movement. An algorithm flags decline or falls, pushing alerts to nurses' phones and medical records. UNN's Centre for Patient-Centred AI (SPKI) runs this vital-sign tracking for safer care and saved nursing time.
NHS England — NHS England's A&E demand forecasting tool now covers 50 NHS organisations, giving bed managers a daily attendance prediction from past admissions and Met Office weather data for staffing and capacity planning. All NHS trusts can use the emergency department attendance prediction system, which had 170 active users by December 2025.
İstanbul İl Sağlık Müdürlüğü (Istanbul Provincial Health Directorate) — Istanbul Provincial Health Directorate now runs an AI newborn screening system citywide, financed with Istanbul Development Agency support. A positive heel-prick result triggers an instant booking at the nearest specialist clinic, and about 150 thousand infant tests get monitored yearly. Autism, cancer and hearing checks join this AI-based screening push next, spanning every age group.
Fundación Clínica Shaio — Fundación Clínica Shaio now leads Colombia as the first hospital applying AI-assisted optical coherence tomography (OCT) in haemodynamics work. Scans reveal fine coronary detail, guiding staff on plaque composition and size ahead of stent placement. Better coronary imaging data means sharper treatment calls.
Taichung Municipal Hospital for Geriatric Rehabilitation and Comprehensive Care — China Medical University Hospital and EverBot Technology built the machines: 20 Aibo units now stationed inside Taichung Municipal Hospital for Geriatric Rehabilitation and Comprehensive Care, reinforcing staff duties as part of Taiwan's growing use of AI-powered hospital robots. Ward navigation for patients, delivering health education material and running telemedicine calls fall to each device, which sharpens its output through ongoing interaction. One shared AI server can coordinate as many as 30 of these healthcare service robots at the same time.
Nguyen Trai Hospital — Doctors in Con Dao's special zone can now flag atrial fibrillation and myocardial infarction through an AI heart rhythm screening tool handed over by Nguyen Trai Hospital in Ho Chi Minh City. Built from Vietnamese patient records rather than borrowed foreign data, the system hits near 100% sensitivity for irregular-heartbeat detection when paired with smart wearable devices. It fills a gap at remote clinics with no cardiologist on staff, extending AI-based cardiac diagnosis beyond major urban centres.
Uganda Ministry of Health — Uganda's Ministry of Health has deployed AI tuberculosis detection through 14 ultra-portable digital X-ray machines fitted to mobile clinic trucks under a US-Uganda partnership with MinXray. The AI-enabled chest X-ray screening system reads scans instantly, flagging TB and signs of up to 20 other lung diseases in communities lacking diagnostic access. The effort targets remote areas previously outside reach of standard screening infrastructure.
Bajcsy-Zsilinszky Hospital Budapest — A Siemens Somatom go top mobile unit now equips Budapest's Bajcsy-Zsilinszky Hospital, overseen by the National Hospital Directorate, in an AI-supported CT scanning rollout combining dose profiling with iterative reconstruction to shrink radiation levels across stroke and general imaging work. Close to 230 people had already gone through it within a fortnight, pointing to fast early uptake in AI-assisted radiology.
Ministerio del Poder Popular para Ciencia y Tecnología — One unified patient file, nationwide: Venezuela's Historia Clínica Única (HCU) gives every public hospital access to an AI-powered health record system running on domestically engineered artificial intelligence. Its built-in assistant helps clinicians move faster through consultations, powering this unified medical record platform as adoption spreads.
Sağlık Bakanlığı (Turkish Ministry of Health) / ASELSAN — Turkish public hospitals now use ASELSAN's HealthView ADR-M100, an AI-powered mobile X-ray built with domestic engineering, producing chest-scan reports within seconds. Needing no internet link, it flags pneumothorax, pneumonia and lung fluid, giving doctors fast AI chest imaging diagnostics at the bedside. The Turkish Ministry of Health has placed 30 units so far, part of a rollout expected to reach about 330 machines.
Ministry of Health / Enmore Regional Hospital — Enmore Regional Hospital in Guyana has adopted an AI medical imaging diagnosis system to read X-rays, CT scans and ultrasounds. Minister Frank Anthony said the AI-assisted radiology tool flags within 14 seconds whether a scan is normal or shows a pathology, producing a preliminary reading radiologists then confirm.
Tribhuvan University Teaching Hospital (TUTH) — ChestGuru, an AI chest X-ray diagnosis tool from Nepali firm Second Eye, is on trial at Tribhuvan University Teaching Hospital in Kathmandu. Scanning one radiograph, it flags 14 possible chest ailments, tuberculosis and pneumonia among them, to speed sorting of urgent cases where wards get crowded.
CHU de Toulouse — CHU de Toulouse, France's first teaching hospital running inHEART, uses this AI 3D heart map for AI heart surgery guidance in tough ablation cases. MRI or scanner scans build the digital twin; hospital figures show timing dropped from five hours to two, success up from 60% to 75%.
Nordsjællands Hospital — Nordsjællands Hospital's emergency department uses AI patient flow forecasting from Praemostro to project arrivals up to 12 hours ahead, hour by hour, letting staff reallocate resources and adjust staffing before demand peaks using this AI arrival prediction in emergency departments. Denmark's Capital Region Centre for IT and Medical Technology worked with Praemostro on the rollout.
First Medical — First Medical, a Jersey GP practice across St Helier, St Peter and St Clement, uses Skin Analytics' DERM app for AI skin cancer detection, which its maker estimates could cut waits, safely discharging up to 50% of cases. Trained on common cancers and benign lesions, this AI-assisted screening assesses mole and lesion photos, speeding clinical decisions and curbing unneeded referrals.
Kyushu University Hospital — Kyushu University Hospital says its AI DPC coding tool from Ubie has added over 65 million yen a year in revenue while shortening staff working hours. Built on a large language model tied into electronic health records, the software scans unstructured clinical notes and recommends diagnosis and procedure codes, matching skilled coders' choices in 90% of cases. The case is a rare example of AI-assisted diagnosis and procedure coding showing a direct, quantified financial return inside a Japanese hospital.
Yegoryevsk Hospital (Moscow Region) — Chest CT and mammography scans at Yegoryevsk Hospital now get automated readings under the Third Opinion service, part of a full AI hospital rollout in the Moscow region. Intensive care staff use video analytics that spot up to 60 adverse event types, while a pilot called ArtInCol assists with colonoscopy. Chief physician Dmitry Kaprin said these tools track patients from diagnosis through recovery, making Yegoryevsk the region's first hospital to apply artificial intelligence across its full care pathway.
Severance Hospital — Y-KNOT, an AI emergency department discharge notes system built by a Yonsei University College of Medicine team, finishes Severance Hospital's mandatory ED paperwork in 32 seconds, down from 69.5 seconds when clinicians handle it unaided. This AI-assisted discharge documentation combines an internal large language model with a compact transformer, operating fully inside the hospital's own network so patient records stay put. Human reviewers scored the machine-drafted text above manually written versions for thoroughness, accuracy, brevity and practical value, per findings in JAMA Network Open.
Santa Casa BH — Santa Casa BH is now the first 100% public SUS hospital to deploy AI clinical decision support, having bought Elsevier's AI suite for physicians. The rollout bundles ClinicalKey for current literature, ClinicalKey AI for clinical questions from validated evidence, and Elsevier OrderSets to unify treatment protocols. As Brazil's busiest SUS provider, this brings AI-assisted medical diagnosis into public healthcare.
Hospital del Salvador — Emergency patients at Hospital del Salvador in Santiago will encounter PREMED Sapiens, a new AI pre-consultation emergency care tool from startup Medical Sapiens, now under trial. Upon arrival, the software reads reported symptoms, sorts cases by urgency and spots ones suited to remote telemedicine follow-up, a design meant to shrink queues and stop people walking out before treatment. This AI-based urgency sorting in hospital emergency departments is backed by Chile's Ministry of Health and Corfo through the Juégatela por la Innovación programme.
Krishna Institute of Medical Sciences (KIMS) — Lifesigns wearables feed KIMS Hospital's AI-based patient monitoring system, the Early Warning System, watching Hyderabad in-patients round the clock. It alerts clinicians the instant readings shift, cutting response delay.
Curaçao Medical Center — Curaçao Medical Center (CMC) leads the Caribbean in AI cancer radiotherapy via MIM Software's Elekta One platform. The gap between a patient's CT scan and treatment start now shrinks, with better shielding for healthy organs during dosing. CMC joins early adopters such as the Netherlands Cancer Institute in radiotherapy treatment planning.
Greek Ministry of Health — Sword Health will engineer Greece's AI triage in healthcare push under a fresh state accord spanning over 10 million people. Referral work and guidance shift onto the existing 1566 line, forming an AI-powered front door to healthcare targeted for Easter 2026, with independent watchdogs, the WHO's Greek office among them, checking progress.
Demerdash Hospital (Ain Shams University) — Cairo's Demerdash Hospital runs Egypt's first centralised AI-powered intensive care management setup, spanning over 50 ICU beds via continuous feeds. Dr Mohamed Ismail of the Egyptian Society of Critical Care and Injuries likens its scope to a full medical team watching every room. Staff cite quicker decisions, fewer mistakes and tighter oversight as the AI ICU monitoring system's chief gains.
Sidra Medicine — Doctors at Sidra Medicine now lean on AI fracture detection paediatric imaging as a daily routine within the Emergency Department and Diagnostic Imaging Unit. Linked directly into the hospital's PACS system, the software scans each X-ray for breaks and dislocations, flagging trouble spots without human prompting, functioning like an extra pair of trained eyes. Clinicians report the AI imaging tool for detecting fractures in children has cut waits and lifted their confidence when reading young patients' scans.
Yashoda Hospitals — Yashoda Hospitals, Qure.ai and AstraZeneca launched an AI-powered Lung Nodule Clinic in Hyderabad, screening 17,000-plus chest X-rays for early lung cancer signs. Radiologists confirmed 960 nodules, 136 high-risk. Automated chest X-ray screening speeds lung disease diagnosis.
St. Luke's University Health Network — St Luke's University Health Network now relies on Epic's Deterioration Index for AI cardiac arrest prediction in every med-surg unit across all 16 of its campuses. Continuous readings of vital signs and lab work stream to a Virtual Response Centre, where nurses act on early warning signs before a patient deteriorates, an approach some call AI in hospital patient monitoring. Cardiac arrests dropped 34% and ICU transfers linked to rapid response calls fell 12%, per the network's own figures, not an independent audit.
Government of Malta (Ministry for Health and Active Ageing) — A Digital Health Strategy 2025-2030 milestone: Malta will bring AI in healthcare to national helpline 1400, sorting incoming callers through an AI-based triage tool. Jo Etienne Abela unveiled the move at a WHO-supported conference held in Malta, noting the line has already logged over 11,000 calls.
Hôpital Fondation Adolphe de Rothschild — Hôpital Fondation Adolphe de Rothschild in Paris runs Tandem Health's AI medical scribe in six departments, after a summer trial of 1,500 consultations. The system writes automated clinical documentation from spoken exchanges, no typing or dictation needed: neurosurgery, neurology, ENT, cardiology, pain management and ophthalmology.
AGEL — AGEL, Central Europe's biggest private healthcare group, now runs AI X-ray analysis in hospitals across Czechia and Slovakia using Carebot. About a million scans get processed yearly, catching pneumonia, fractures, effusions and 60-plus other conditions. Seven tools cover bone, lung, brain and mammography across X-ray and CT scans. Coverage began at four Slovak sites plus Nový Jičín, then reached Šternberk, Ostrava-Vítkovice and Skalica, moving toward full AI-assisted radiology imaging group-wide.
Shamiri Institute — shamiriAI debuted at Kenya's Shamiri Institute during the 2025 Shamiri Summit in Nairobi, bringing an AI mental health platform to youth across Africa. Live analysis of counselling talks flags symptom trends and steers users toward what helps, spanning English, Swahili and Sheng via speech recognition, analytics and machine learning. Testing continues in Kenya, ahead of a drive for AI-powered mental health support to reach 10 million youth by 2032.
ASL TO4 - Ospedale di Ivrea — ASL TO4's Ivrea hospital activated a new AI-based imaging platform on 11 November 2025 to improve prostate cancer MRI diagnosis. Suspicious tissue can now be flagged with far greater accuracy thanks to three-dimensional mapping and detection algorithms, an AI-driven diagnostic system that shortens the wait for results and sharpens how radiotherapy gets planned. Funding came from a grassroots campaign run by local service clubs.
Naples Comprehensive Health (NCH) — A United States milestone: Rooney Heart Institute has begun running Caristo Diagnostics' paired tools, CaRi-Plaque and CaRi-Heart, marking the country's first hospital deployment combining both for AI heart attack risk prediction. Reading ordinary CT scans, the pairing warns of trouble up to 10 years before any symptom emerges — FDA-cleared CaRi-Plaque flags dangerous plaque via coronary CT angiography, while its partner tool catches inflammation building even earlier, ahead of visible plaque. The institute is folding this cardiac risk assessment technology into a self-pay preventive heart programme as well.
Larissa General Hospital — Emergency department delays at Larissa General Hospital have fallen by 30 to 40 per cent since clinicians began relying on StapTep, an AI triage system designed to steer arriving patients toward suitable treatment and ease clerical strain on physicians. Greece's Minister of Health cited the outcome as proof that automated patient routing delivers results in practice, and officials now plan to extend the same digital assistant across hospitals nationwide. Speed to treatment, ministry figures suggest, now hinges on clinical urgency rather than paperwork.
Greek Ministry of Health — Greece's Ministry of Health and Ministry of Digital Governance deployed the Digital Doctor Assistant, the first AI in healthcare tool nationally. Inside the National Electronic Health Record (EHR) via myHealthDoc, this medical AI fetches test results, patient histories and medication records from doctors' questions, trimming paperwork and backing clinical decision-making.
Hospital Municipal de Agudos Leónidas Lucero de Bahía Blanca / ICIC (CONICET-UNS) — AI patient monitoring at Hospital Municipal de Bahía Blanca now uses a tool from ICIC (CONICET-UNS) to scan records for multiple chronic conditions, combining natural language processing with ICD-10 coding. It maps risk groups for AI chronic disease detection, guiding personalised treatment. Staff were trained, and its code is open access.
Schneider Children's Medical Center of Israel (Clalit) — Paediatricians at Schneider Children's Medical Center in Israel now query PANDA (Pediatric AI Navigation and Decision Assistant) in plain language and get instant, protocol-grounded answers as part of an AI clinical decision support pilot. Built jointly by Clalit Innovation and the hospital, the tool uses RAG (retrieval-augmented generation) so every response draws only on the centre's own library of over 600 clinical protocols, never outside sources. Clalit intends to extend this clinical decision-making assistant for paediatricians beyond the trial to its full network of 14 hospitals and 1,600 clinics.
CHU de Montpellier — Doctors at CHU de Montpellier can now destroy liver tumours once thought unreachable, using a newly opened Angio-CT suite built around Canon Medical Systems' Alphenix 4D CT scanner, the first AI in interventional radiology installation anywhere outside Japan. Deep learning refines live imaging by fusing angiography, ultrasound and CT scanning into a single space. The radiology AI deployment marks a global milestone for the hospital.
Queen Mary Hospital — Dr Carmen Chan says Queen Mary Hospital will widen chest pain services on 29 October 2025, folding AI ECG triage into heart attack care to speed diagnosis. Her team's model sorts arrivals by low, moderate or high risk, and the highest-risk group gets routed straight to the on-call cardiologist. Since opening, the centre has already managed 1,400 acute cases in electrocardiography-based patient assessment.
Seoul St Mary's Hospital — A new AI voice recognition medical records system, CMC GenNote, is being trialled at Seoul St Mary's Hospital, built with PuzzleAI on LLM-based large language model technology. Rather than simple dictation, the tool lets doctors summon documentation templates and dictate entries directly into them, replacing the transcription-only setup in place since 2019. Currently limited to certain outpatient clinics, the rollout is set to reach surgical suites, emergency departments and general wards next, marking a step toward AI-powered clinical documentation by voice.
Qscan Group — Qscan Group wired its clinics into CARPL.ai, unlocking one shared screen where radiologists tap over 250 AI models for radiology reporting. Called the biggest AI diagnostic tools marketplace anywhere, the rollout covers all of Australia and folds in support made for the National Lung Cancer Screening Program.
Centro de Ingeniería para la Investigación, Desarrollo e Innovación Tecnológica (CIDIT) — Doctors, scientists and pharma firms can now use two biomedical AI tools, CoFar and ProteIA, free of charge, courtesy of CIDIT, a Paraguayan research centre. CoFar predicts likely clashes and side effects when drugs are mixed, whereas ProteIA employs machine learning on protein interactions to further disease and treatment research — part of a wider AI-based biomedical research push.
Uchida Hospital (medical corporation Taiseikai) — Uchida Hospital, Gunma Prefecture, is Japan's first chronic-care hospital using OPTiM AI Hospital for generative AI clinical documentation. The tool drafts doctor's opinion papers, nursing summaries and medical information forms from medical records, plus voice-recorder AI summarisation and beta charting-support and informed-consent summarising. Staff report substantially better, faster automated hospital paperwork, freeing patient-care time.
British Antarctic Survey — The British Antarctic Survey (BAS) has deployed AI seafloor image analysis, cutting per-photo time from up to eight hours to a few seconds. Built by the BAS AI Lab from 100 annotated Weddell Sea images, the tool spots fish, sponges, corals and starfish in video and high-resolution photographs. More than 30,000 backlogged images from the Weddell Sea and Antarctic Peninsula are being processed, and this automated species detection runs in real time aboard research vessels.
Heilbrigðisstofnun Norðurlands (HSN) — Health centres under Healthcare Institution of North Iceland (HSN) now handle AI eye screening for diabetes locally, sparing patients the Reykjavík trip. RetinaRisk, the Icelandic firm behind the tool, projects nearly 99% lower screening costs and hundreds fewer annual journeys. Healthcare Institution of Suðurnes (HSS) previously ran this diabetic retinopathy screening system before HSN's rollout.
Premier Integrated Labs — Mindray's CAL 8000 now runs as Malaysia's first AI-assisted haematology system in private practice, deployed by Premier Integrated Labs. Deputy Health Minister YB Dato Lukanisman Bin Awang Sauni launched the automated haematology analyser, which PIL says sharpens accuracy, speeds turnaround and lifts overall lab efficiency for clinicians reading results.
Hackensack Meridian Health — AI agents in healthcare now cover 18 hospitals and 500 care sites at Hackensack Meridian Health, New Jersey's biggest network, running on Google Cloud's Gemini 2.5. Its staff led the field by adopting a Gemini note-summarising tool at scale, with over 7,000 clinicians producing more than 17,000 summaries since June 2025. This clinical AI deployment further includes a NICU nurse agent surfacing best-practice guidance and a lab values agent spotting trends to suggest preventive steps for primary care doctors.
Central Adelaide Local Health Network (CALHN) — 500 patients at The Queen Elizabeth Hospital and Sefton Park Health are trialling CALHN's AI medical scribe in SA emergency rooms. A compact microphone converts spoken exams into finished notes instantly, easing staff workloads. SA clinicians built this AI scribe technology for SA Health, so recordings stay on SA Health's servers.
Ministry of Health, Kenya — Kenya's Ministry of Health is scaling up AI TB detection: 80 ultra-portable digital X-ray machines now reach 43 counties. Aden Duale unveiled them at KEMSA's Nairobi base, running Computer-Aided Detection software that reads scans unaided, catching early tuberculosis and chest illness signs. Priority goes to slum dwellers, inmates and miners under this AI-powered chest X-ray screening push.
Sophiahemmet Husläkarmottagningen — Sophiahemmet's Stockholm clinic, Husläkarmottagningen, now runs Tandem, an AI medical documentation tool acting as a virtual secretary. It overhears consultations and auto-generates referrals, certificates and notes, easing paperwork for more patient time. This marks wider AI clinical documentation use in Sweden's primary care.
Regency Specialist Hospital (HMI Medical) — Johor gains its first AI smart ward as Regency Specialist Hospital activates the Next-Gen Smart Ward, pairing bedside terminals with a nursing dashboard and automated vital-signs capture to ease pressure on nursing staff. Parent group HMI Medical says manual data entry now takes 67% less time, alongside a 16% quicker response rate, an AI-enabled hospital ward result the hospital itself has not had independently audited.
Health New Zealand — Ask Groov, an AI therapy chatbot funded by Health New Zealand Te Whatu Ora (Health NZ), gives young people round-the-clock mental health advice. Sir John Kirwan's firm Groov trained its generative AI on guidance from certified psychologists and psychiatrists, skipping raw web data. The aim: pull young people from unregulated AI mental health support toward something screened.
HealthBay Vitalia — Dubai's HealthBay Vitalia clinic runs the UAE's first Hologic 3Dimensions system with Genius AI Detection for AI mammography and AI breast cancer detection. Software scanning every image layer hits 94 per cent sensitivity, cutting needless callbacks by up to 70 per cent versus older machines. Doctors catch 20 to 65 per cent more invasive tumours, especially in patients with dense breast tissue.
Great Ormond Street Hospital for Children NHS Foundation Trust — Every outpatient clinic at Great Ormond Street Hospital (GOSH) has switched on AI scribe capability, letting TORTUS listen through ambient voice capture and auto-generate clinical notes. Earlier NHS research spanning nine sites and 17,000-plus interactions recorded a 23.5% boost in direct interaction time plus 8.2% shorter appointments.
Ministerio de Salud de El Salvador (Ministry of Health) — El Salvador's public hospitals now perform AI X-ray analysis in hospitals on every radiograph and lab test, Minister Francisco Alabi says. This AI-assisted radiograph screening flags abnormalities faster, catching respiratory, cardiac and oncological problems sooner. Doctors still make the final call — the tool aids diagnosis, not replaces it.
Pyongyang General Hospital — North Korea unveiled its AI smart medical system at the newly opened Pyongyang General Hospital, an intelligent medical service system built on smart software. Kim Jong Un attended and pressed for constant upgrades, wanting the AI-driven setup to sharpen patient care and hospital administration.
Oy Apotti Ab — HUS, Helsinki and Vantaa-Kerava are trialling AI patient record summarisation within Apotti, their shared health record platform. Staff across primary care, specialised medicine and social services get condensed views of a person's history instead of full files, via AI-generated clinical summaries meant to speed up how carers grasp each case.
Kazakhstan Ministry of Healthcare — 30 clinics scattered across Kazakhstan now form the testing ground for an AI therapist, unveiled amid Digital Bridge 2025 forum proceedings. Officials envisage nationwide adoption across every medical organisation once this trial mental health chatbot proves itself. Broader digital healthcare projects already underway in the country provide the backdrop for this initiative.
Dr Lal PathLabs — India's Dr Lal PathLabs now leads in AI-assisted cancer diagnosis, becoming the first domestic lab to deploy a deep learning module flagging lymph node metastasis, including micrometastasis. Built with Qritive, the QiAI Lymph Node Dx tool caught every true positive across tested cancer types, with zero false negatives recorded. Pathology AI screening like this now sits inside the lab's diagnostic workflow.
Kyrgyz scientists at National Center for Cardiology and Therapy — Kyrgyzstan's National Academy of Sciences trialled an AI heart failure detection system at a Bishkek care home, matching ultrasound findings 85 per cent of the time. The AI cardiac diagnostic tool reads electrocardiogram frequency patterns to flag hidden decline within five minutes. Researchers want the Ministry of Health to back wider rollout.
Matsunami General Hospital — AI DPC coding is now live at Matsunami General Hospital in Gifu, linked to electronic health records via Ubie's generative AI platform since August 2025. Messy clinical notes become organised data, tightening diagnosis-procedure combination accuracy. Automated discharge summaries are planned next.
Government of Guyana / Ministry of Health — Cutting drug spoilage and lifting region-wide efficiency: that's the goal behind Guyana's Materials Management Unit adopting AI-driven automation for medical supply chain management from the first quarter of 2026, handling AI-driven pharmaceutical distribution nationwide. President Dr Mohamed Irfaan Ali tied the plan to broader AI technologies goals for the country's healthcare system.
Tianjin Medical University General Hospital — Nearly 100% analysis accuracy and 49.3% faster reports: that's the claim behind an AI in hospital diagnostics tool from China Mobile Tianjin and Tianjin Medical University General Hospital, built on DeepSeek-V3. Deployed at the check-up centre, this medical health management large model drafts health check reports automatically, aids clinical decisions and reads patient records and scans. The developers, not an independent auditor, supplied these figures.
Azienda Ospedaliera Sant'Andrea — Sant'Andrea University Hospital in Rome now uses AI-powered digital pathology for cancer diagnosis, Lazio's first full deployment of its kind. One linked system pairs GE HealthCare's Enterprise Imaging with Tribun Health's CaloPix, pulling radiology, pathology, cardiology and nuclear medicine data into AI-assisted diagnostic imaging. Nearly 1.5 million euros funded the build, partly via Italy's PNRR plan.
Dirección de Redes Integradas de Salud (Diris) Lima Centro — Five portable AI X-ray TB diagnosis machines are now running across Lima clinics under Diris Lima Centro, Peru's health network, aiming to catch tuberculosis sooner. This AI-assisted tuberculosis screening reads each scan in 2 to 3 minutes, spotting fibrosis, nodules and other lung abnormalities alongside TB. Global Fund cash backs the rollout, live at five locations so far, among them San Juan de Lurigancho, site of 1,270 TB cases in 2025.
Ethiopian Ministry of Health — Ethiopia's Ministry of Health has deployed 195 machines for AI-powered X-ray TB screening nationwide, backed by a 2.6 billion birr investment across regional health facilities. Some 250 health workers were trained to run the AI tuberculosis detection equipment, which reads chest radiographs and flags likely cases for faster diagnosis and treatment.
King Faisal Specialist Hospital & Research Centre — Over 30 AI healthcare applications now run at King Faisal Specialist Hospital & Research Centre (KFSHRC), plus five new AI agents added this year, spanning clinical documentation, medical translation, data analysis, chest X-ray and brain CT radiology tools, and live patient experience monitoring. Its Centre for Healthcare Intelligence has separately added over 20 AI tools for diagnostics, patient flow and predictive care, part of a broader AI in hospitals drive across Saudi Arabia.
MedIQ Smart Healthcare — MedIQ Smart Healthcare debuted its AI Suite at the Med-AI Health Summit in Islamabad, billed as the nation's first AI-powered assistant serving both physicians and patients. The system pairs an Urdu chatbot with symptom-tracking tools that keep medical history on file and relay it to providers, then points users toward specialists and dosage advice, a clear step forward for AI in healthcare. It stands out as an early digital health assistant tailored to Pakistan.
Cleveland Clinic — A 46% jump in caught sepsis cases at one pilot site, plus a tenfold cut in false alerts, is why Cleveland Clinic is widening its AI sepsis detection push nationwide. Bayesian Health's clinical intelligence platform already covers thirteen hospitals, with Ohio and Florida joining through 2025. Clinicians gain faster onset warning, spotting risk patterns older alerting tools missed.
Hanoi Hospital of Post and Communication (Bệnh viện Bưu điện) — Hanoi Hospital of Post and Communication is testing EIRL, a Japanese AI diagnostic imaging tool from VNPT-IT, LPIXEL and JTEC, through 28 February 2026, scanning chest X-rays, CT and brain MRA. Japan runs this radiology AI in 1,000-plus hospitals across 12 million cases. The pilot tests whether that record travels.
University Hospital in Kraków — AGH University of Science and Technology's Raygenic-built Rayspad platform completed six months of clinical trials at Kraków's University Hospital, outlining organs, flagging abnormalities including possible tumours, and measuring CT and MRI scans. Radiologists called the AI diagnostic tool an unusually thorough assistant for spotting malignancies in medical imaging.
University Hospital Kanev — AI radiology arrives in northern Bulgaria: University Hospital "Kanev" in Ruse now runs a certified European system from French firm AZmed, flagging suspect X-ray findings for staff review. Automated X-ray screening catches fine fractures a busy eye might miss, giving emergency trauma cases a built-in double-check. The tool assists radiologists rather than replacing them.
NYU Abu Dhabi, Yas Clinic and Abu Dhabi Stem Cells Centre — An AI-powered MRI technique for quicker multiple sclerosis diagnosis in Abu Dhabi is entering clinical validation at Yas Clinic-Khalifa City. NYUAD (New York University Abu Dhabi) and the Abu Dhabi Stem Cells Centre built the FLAIRstar module, which automatically flags the central vein sign, a lesion marker unique to MS, enabling AI-assisted MRI diagnosis without manual review. Doctors trialling it report shorter waits and earlier treatment.
Queen Alia Heart Institute (Royal Medical Services) — An AI-integrated ICU monitoring system, the ARC device, now runs in the Queen Alia Heart Institute's open-heart surgery unit under Jordan's Royal Medical Services — a regional first cleared by the US Food and Drug Administration. Wired into ventilators, infusion pumps, dialysis machines and electronic medical records, it merges bedside readings into one picture for faster clinical response. Officials call it a milestone for AI in intensive care, meant to sharpen urgent decisions after staff training.
Hospital General Riobamba (IESS) — IESS, Ecuador's social security institute, has rolled out IA PROA at Hospital General Riobamba to flag, refine and track restricted antimicrobial use. The hospital calls it Latin America's first AI antibiotic stewardship system, an AI-driven antimicrobial resistance monitoring tool overseeing rational prescribing of broad-spectrum antibiotics.
Kameda General Hospital — Kameda General Hospital in Chiba Prefecture now lets physicians pull up records by speaking or typing plain questions, via Ubie's generative AI Patient Navigator and AI Partner tools for AI clinical decision support. Scattered files across electronic health record and departmental systems merge into one lookup point, sparing staff the mental strain of hunting for data and backing diagnostic decision-making.
Social Health Authority (SHA) — Kenya's Social Health Authority plans to deploy an AI clinical support tool prescriptions can be cross-checked against, matching medications to patient records on the Afya Yangu portal and automatically blocking drugs likely to cause harm such as allergic reactions. SHA CEO Mercy Mwangangi described the AI-powered prescription safety system on September 16, 2025, with the Ministry of Health expected to launch the technology in 2026.
Volta Redonda municipal government — Volta Redonda has become Rio de Janeiro state's first municipality to apply AI for imaging diagnostics, with São João Batista Hospital, Dr Nelson dos Santos Gonçalves Hospital and the Santo Agostinho emergency unit now feeding chest X-rays to HarpiAI for review. An automated radiology second opinion arrives in under five minutes, ahead of a planned expansion into mammography. Doctors keep final authority.
Novosibirsk Region Health Ministry — Roughly 250,000 scans have run through AI radiology in Novosibirsk region's hospitals since January 2025, covering mammograms, CT scans and fluorography images. The Novosibirsk Regional Clinical Hospital and other facilities pair the Third Opinion and MosMedII platforms with the Central Archive of Medical Images (TsAMI) for this AI-assisted medical imaging. A stroke-detection brain CT module, added August 2025, resolved two of three unclear cases within 20 minutes.
John Paul II Specialist Hospital in Cracow — Cracow's John Paul II Specialist Hospital now runs BrainScan S.A.'s brain CT AI tool for reading scans. It marks abnormal patterns itself, giving doctors faster grounds for treatment calls. This brings AI-assisted brain scan interpretation into daily use at a top medical centre in Poland.
Labaid Group — Labaid Group debuted two AI healthcare tools at Labaid Cancer Hospital and Super Speciality Centre in Dhaka: Labaid-GPT, giving physicians rapid, evidence-based recommendations from treatment guidelines and clinical records, and LUNA, an AI-driven clinical support chatbot cycling through eight functions from nursing to legal counsel. Eight patents protect the pair, unveiled 7 September 2025.
Santé publique France — France's public health agency tests Mistral AI's generative model for AI disease surveillance, processing large datasets and speeding epidemiological summaries. Santé publique France says this generative AI for epidemiological analysis speeds analytical programme development while meeting data security rules. It has also begun building its own high-performance computing infrastructure.
Bangladesh Medical University (BMU) Super Specialized Hospital — A new robotic rehabilitation centre, Bangladesh's first, has opened at Bangladesh Medical University (BMU) Super Specialized Hospital, cramming 57 machines, 22 of them AI-driven, into 8,000 square feet below ground. Stroke-related paralysis and long-term nerve disorders now fall within reach of trial care there, scaling toward 50 people served each day at full tilt. Nurjahan Begum cut the ribbon, backed by Chinese government money behind the AI rehabilitation robotics push.
Hospital Carlos Van Buren — Chile counts merely 47 geneticists serving over 19 million residents, prompting Hospital Carlos Van Buren in Valparaíso to trial DIAGEN-IA, an AI platform for paediatric genetic diagnosis built by Universidad Andrés Bello's ITiSB institute toward quicker paediatric genetic disease detection. Symptom entries yield ranked possible conditions with suggested tests. Nicole Nakousi heads the effort alongside neuropediatric staff.
Egypt Healthcare Authority (EHA) at Suez Medical Complex — AI in colonoscopy has arrived in Egypt, deployed at the Suez Medical Complex by the Egypt Healthcare Authority. The system gives doctors a live diagnostic aid during exams, scanning images to spot small tumours early and guide precise treatment. A Gastroenterology Scientific Day also featured other advanced work — bile duct stone removal, oesophageal myotomy and ultrasound-guided pancreatic biopsies — part of this AI-powered gastrointestinal endoscopy push.
Peking University Shenzhen Hospital — An example of AI in hospital patient services: Beidafu, a virtual assistant at Peking University Shenzhen Hospital running on Ant Group's Bailing medical LLM. This hospital patient AI assistant guides patients through 27 checkpoints from arrival to result collection, pushing alerts on what comes next instead of waiting to be asked, ending the hospital's old passive setup.
UM Specialist Centre (UMSC) — UM Specialist Centre has switched on Q-Ward, Malaysia's first fully operational AI-driven smart ward, fitting bedside screens with artificial intelligence for routine surgery patients. One panel per bed covers dining, entertainment, charts and lab results together. UMSC says the AI hospital ward helps offset staffing gaps by giving carers live case insight.
Public Hospitals Authority — The Bahamas' Public Hospitals Authority (PHA) began an eight-week AI senior monitoring trial, Remember 2 (RM2.ai) Pilot Program, fitting around 100 older adults from the Geriatric Hospital and affiliated clinics with app-linked smartwatches. AI-driven elderly monitoring of vital signs, medication reminders and fall prevention aims to cut hospitalisations.
Royal Brisbane and Women's Hospital — Annalise.ai, AI chest x-ray analysis software able to flag as many as 124 possible findings, is being piloted at Royal Brisbane and Women's Hospital for six months. Australia's public sector sees its first hospital test of the tool's secondary capture function here, using AI-powered medical imaging software as decision support rather than replacing clinical judgement. Final sign-off still rests with staff reviewing each scan before reports are issued.
Vanderbilt University Medical Center — About 1,300 clinicians use DAX Copilot, Microsoft's ambient documentation assistant, after Vanderbilt University Medical Center widened access beyond an initial trial. Emergency and outpatient departments deploy the AI scribe for clinicians to capture visits and draft notes automatically. Vanderbilt next plans ambient AI notetaking for hospital wards and nursing teams.
Ministry of Communications and Information Technology / Baheya Foundation — Egypt's Ministry of Communications and Information Technology built a locally engineered AI breast cancer detection system, now live at Baheya Hospital's Sheikh Zayed and Haram sites. It reads mammogram scans for likely tumours at roughly 90 per cent accuracy, a developer-reported figure, not independently audited. The aim: faster breast cancer screening AI diagnosis and better recovery odds.
Oncoclínicas and Casa de Saúde São José — A Brazilian AI lung cancer detection tool from Oncoclínicas and Casa de Saúde São José flags hidden nodules in every chest CT in Rio de Janeiro, whatever the exam's original reason. Twelve months, 6 mil scans, 150 nodules caught early, three tumours confirmed, turnaround under 24 hours. Salvador and Brasília are next for the AI cancer screening effort.
Health New Zealand (Te Whatu Ora) — Nicolab, a Dutch AI firm, is bringing its AI stroke imaging platform StrokeViewer into up to 30 New Zealand hospitals under a Health New Zealand rollout backed by the University of Otago. A Health Research Council-funded trial will check if faster scan reading cuts stroke treatment delays, targeting 850 additional patients reaching reperfusion therapy yearly. It marks a large-scale test of AI-assisted stroke diagnosis, pending confirmed outcomes.
Halodoc — AIDA, Halodoc's AI medical assistant for partner physicians, hit 93% accuracy handling administrative work during its first rollout. A board of medical excellence guided its doctor-first design, freeing clinicians for diagnosis and treatment decisions instead of paperwork. Over 80% using this AI Doctor Assistant reported clear benefits.
Health Plus — Health Plus has become the first Jersey GP practice to use an AI scribe, adopting Heidi Health to transcribe doctor–patient consultations. Backed by Digital Jersey, the software drafts patient summaries, medical notes and referral letters automatically, so GPs spend less time typing and more time with patients.
National Tuberculosis Programme of the Central African Republic, The Union and EPCON — An AI model for TB hotspot mapping split Bangui into 100-metre grids, surfacing clusters the Central African Republic's National Tuberculosis Programme (NTP) never knew existed. Built with The Union and EPCON using density and clinic-distance data, findings published 13 August 2025. Officials then targeted screening at districts with the heaviest TB burden.
Butabika National Referral Mental Hospital — Hard-to-reach Ugandan communities can now dial into care via a new AI mental health call centre, built jointly by Butabika National Referral Mental Hospital, Makerere University and Tanzania's Mirembe Mental Health Hospital. Stretched clinicians gain breathing room, as AI-powered telehealth counselling extends scarce staffing further.
Ministry of Health, Eswatini — Eswatini's Ministry of Health, with the Taiwan Technical Mission, has installed AI eye screening equipment for diabetic retinopathy at six sites: Mbabane Government Hospital, Piggs Peak Hospital, Mankayane Hospital, Raleigh Fitkin Memorial Hospital, Dvokolwako Health Centre and Emkhuzweni Health Centre. The AI-powered diagnostic system for diabetic retinopathy flags diabetes-related retinal damage, speeding diagnosis to cut preventable blindness in rural areas.
Kingston Health Sciences Centre — Canada's first hospital use of HeartFlow puts AI coronary artery disease diagnosis inside Kingston Health Sciences Centre, turning CT scans into 3D arterial maps that calculate blood flow and avoid catheterisation. Stephen Sorensen's $100,000 gift paid for it. Doctors now have a non-surgical path to AI heart disease detection in blocked arteries.
ASST Brianza — Real-time AI X-ray reading now supports emergency departments at three ASST Brianza sites: Carate Brianza, Desio and Vimercate. Scans are interpreted as captured, narrowing diagnostic error and shortening patient waits. Officials call it among Italy's earliest routine clinical AI rollouts, also applying AI-assisted radiology diagnosis methods to colorectal cancer screening in gastroenterology.
Krajská zdravotní — Real-time flagging of suspected healthcare-associated infections is now handled by AI hospital infection surveillance at Krajská zdravotní, the operator behind a network of seven hospitals in Czechia's Ústí nad Labem region. Rather than relying on staff to submit reports after the fact, the system continuously scans electronic patient records and alerts infection control teams the moment a case looks likely. Ward-by-ward epidemiological trends over time are now visualised graphically, giving clinicians a picture that the prior passive reporting model never offered — a shift industry watchers might call automated healthcare-associated infection monitoring.
Hsinchu Mackay Memorial Hospital — Hsinchu MacKay AiCare runs entirely on hospital servers, a design choice by Phison Electronics and Hsinchu Mackay Memorial Hospital meant to stop patient data leaving the building. The generative AI system supports doctors through record assistance, translates for patients before surgery in multiple languages, manages medical knowledge, summarises meeting audio and answers health questions via chatbot, illustrating one model for AI in hospital care. Hsinchu City Mackay Children's Hospital has since adopted the same setup, extending this healthcare AI rollout.
Conselleria de Sanidad de la Generalitat Valenciana — AI X-ray analysis now covers every hospital and primary care site in Spain's Valencian Community, spanning chest, bone and breast imaging, per Marciano Gómez. Over 300,000 chest and bone studies plus roughly 100,000 mammograms have gone through the software, delivering results within five minutes at a negative predictive value nearing 100%. The AI radiology diagnostics tool aids clinicians without overriding judgement.
Forte Diagnostics — Forte Diagnostics debuts an AI-powered medical diagnostics platform, Sri Lanka's first, made with India's Health Vectors. Bland text reports become live, personalised dashboards under this health data visualisation model.
Cheshire and Merseyside Radiology Imaging Network — Cheshire and Merseyside Radiology Imaging Network (CAMRIN) has switched on Annalise.ai's AI for lung cancer detection at all 11 imaging sites. Chest X-rays are screened for possible tumour signs, with radiologists alerted in under a minute. NHS AI Diagnostic Fund backed the £1.2 million project, an AI-assisted chest X-ray screening scheme now shortlisted for a Health Service Journal Digital Award.
Presidio Ospedaliero di Tolmezzo — Tolmezzo hospital in Friuli, Italy, has launched the nation's first radiology trial for AI fracture detection radiology, using Galileo, built by General Medical Merate and Milvue for AI-assisted bone imaging. The system flags limb breaks, certain lung and pleural problems, works out spine, pelvis and lower-limb dimensions unaided, and predicts children's bone age. The University of Campania handles the trial's evaluation.
Government of Aragón (Servicio Aragonés de Salud) — Servicio Aragonés de Salud now runs Salud-GPT, Aragón's Government's first corporate AI diagnosis tool for health professionals. Workers enter symptoms and get back conditions ranked by likelihood, built to spot rare diseases sooner via the electronic health record. Joaquín Velilla, Digital Health director general, calls this AI clinical decision support tool a guide, not a replacement, for professional judgement.
亞東醫院, 中山醫學大學附設醫院, 高雄榮民總醫院, 奇美醫院 — Four Taiwanese hospitals — Far Eastern Memorial Hospital, Chung Shan Medical University Hospital, Kaohsiung Veterans General Hospital and Chi Mei Medical Center — have adopted MedBobi 2.0, a generative AI platform built by the Industrial Technology Research Institute, bringing AI clinical documentation into everyday practice. Built on voice recognition paired with patient records, the system pushes clinicians beyond routine paperwork into faster diagnostic reasoning, shrinking the write-up and evaluation stage for each case from 30 minutes down to 3 minutes. That gain comes from ITRI's own internal measurement of AI-assisted clinical decision support, not an outside audit.
Medical Specialist Group (MSG) — Ten consultants trialled Heidi, an AI medical scribe now spreading across every outpatient clinic at Guernsey's Medical Specialist Group after a June 2025 pilot, part of a wider shift toward AI-assisted clinical documentation. It listens throughout appointments, generating referral letters and summaries, though staff must still verify and sign each output before it reaches patients.
Grodno city clinical hospital of emergency medical care — A Grodno hospital's AI assistant emergency department tool now poses over 30 pre-medical screening questions to arrivals, passing replies straight to the attending physician. Built with TeleMed AI laboratory, the launch adds a digital assistants' office and smart queue kiosk. This automated patient-triage technology aims to shorten waits before clinicians see patients.
Assistance Publique - Hôpitaux de Marseille (AP-HM) — AP-HM's Hôpital de la Timone in Marseille now runs GE HealthCare's CleaRecon DL, the second global site and first in Europe using AI in interventional radiology. The deep-learning core sharpens CBCT scans, letting clinicians trace finer arteries through AI-based 3D image reconstruction for safer, precise treatment, including tumour embolisation.
University of the Free State and Universitas Academic Hospital — MD Anderson Cancer Center's Radiation Planning Assistant debuted clinically at Bloemfontein's Universitas Academic Hospital and University of the Free State, a world first for AI cancer radiotherapy planning. Cervical cancer cases opened the work, later joined by breast, head, neck and brain tumours, lifting automated radiotherapy planning totals near 50 patients.
Bruyère Health — Bruyère Health fitted Esprit-ai's Sense platform in its Ottawa home, sensors under mattresses catching bed exits, lowering fall risk in long-term care. A nine-month trial came first: 91% alert adherence, zero falls logged. Alerts now flag pressure injuries early, reducing falls in care homes.
Hallym University Medical Center — Doctors at Hallym University Medical Center will gain back 20 to 30 working days each year, thanks to a generative AI platform called HAI (Hallym Artificial Intelligence) that now handles AI medical records duties across the whole hospital stay, from a patient's arrival to their release. Built in partnership with Konan Technology, HAI pulls together 97 distinct categories of clinical information and is forecast to eliminate about 83,000 hours of manual charting work per year. So far it handles five conditions - stroke, gallbladder inflammation, caesarean delivery, cataract procedures and tonsil removal - operating on a locally hosted LLM rather than a cloud service, which keeps sensitive health information from leaving the hospital, an approach that supports broader AI clinical documentation goals in hospital settings.
Manipal Hospitals — Manipal Hospitals runs AI nurse handoffs at 23 of its 37 sites, cutting handover time from 90 to 20 minutes — a 70% drop. Google Cloud built it on Gemini on Vertex AI; over 5,000 nurses use it daily. The Indian chain's shift-transfer tool anchors clinical communication.
Ministry of Health and Social Action of Senegal and Kajou — No signal, no problem: kSANTÉ fits AI e-learning for community health workers onto one MicroSD card across Senegal. Kajou joined the Ministry of Health and Social Action to load 14 lessons plus an AI-powered training platform for health workers, drawing on official archives, onto each card. Dakar launched with 2,000 units.
Meuhedet Health Fund — Israel's Meuhedet Health Fund now runs Meuhedet AI, an AI assistant in healthcare app form, an AI-powered patient support tool inside its member portal. Members type everyday requests to locate medical information, book services or fetch documents, avoiding the older menu structure. It builds on Meuhedet's prior Noach chatbot, applying natural language processing to read intent, and functions across every kosher-filter-compatible device.
Hull University Teaching Hospitals NHS Trust — AI MRI scan times fell 10-15 minutes at Hull Royal Infirmary and Castle Hill Hospital after Hull University Teaching Hospitals NHS Trust fitted Air Recon Deep Learning software to its scanners. This image reconstruction technology strips noise, cutting a head scan from 30 to 20 minutes. Lumbar spine slots per shift rose from 21 to 31.
PATH and Viamo — CHEWA (Community Health Extension Worker Assistant), an AI health hotline in Nigeria devised by PATH and Viamo, needs nothing but a basic handset to reach frontline staff. Callers in Kano and Oyo states receive spoken, generative-AI clinical advice, an offline-friendly example of voice-based AI clinical decision support built for places lacking data access. Around 100 participants signed on, enrolment launched June 2025, with a goal near 3,000 patient encounters.
Kassenärztliche Vereinigung Schleswig-Holstein — 500,000 euros from Schleswig-Holstein's health ministry now backs AI mammogram screening within QuaMaDi (Qualitätsgesicherte Mamma-Diagnostik), Germany's breast diagnostics scheme. The tool highlights irregular tissue patterns and scores malignancy likelihood, aiding radiologists' double-reads for higher-risk women. Lübeck University will assess this AI breast cancer detection approach.
Khanty-Mansi Autonomous Okrug Department of Health — Reading times dropped 58% after Nikolay Pirogov, a floating polyclinic under Khanty-Mansi Autonomous Okrug's health department, adopted AI X-ray diagnostics. Third Opinion AI, from the Medical Information and Analytical Centre, drafts findings as the ship reaches remote Ob and Irtysh river villages. A breast-screening AI imaging analysis tool follows July 2025.
Stavanger Municipality Emergency Room — AI patient inflow prediction now runs at Stavanger municipality's emergency room, forecasting arrivals hours and days out via a model built on Facebook's Prophet and Microsoft Fabric. Historical attendance figures feed the forecast, helping planners set staffing levels and shrink queues. What began as a proof of concept has since reached partial deployment, sharpening AI patient flow prediction for clinicians and the public alike.
Hospital Clínico San Carlos — A global first: Hospital Clínico San Carlos, a public hospital in Madrid, steered a coronary catheter via augmented reality and AI into an 89-year-old with several clogged arteries. AI in percutaneous coronary interventions blends AI-read CT scans with live X-rays mid-operation — AI-assisted cardiac catheterisation merging two imaging streams at once.
Tampa General Hospital — Tampa General Hospital is putting AI in nursing documentation to work, wiring Microsoft's ambient listening into Epic's Rover mobile app so bedside chats turn into finished charting within seconds. Nurses gain back time once lost to typing, mirroring the ambient AI for healthcare documentation already given to more than 500 physicians via DAX Copilot since June 2024 — only now reaching the nursing floor.
Medentec — Medentec debuted Meri Muskan, Pakistan's first AI dental telehealth platform, from Karachi's National Incubation Center. A homegrown model reads uploaded photos and symptom notes to flag oral trouble, then links users to licensed dentists over encrypted video. Prescriptions, records and delivery round out Karachi service, extending dental care access powered by AI.
Karolinska University Hospital — AI surgery scheduling has arrived at Karolinska University Hospital, where a new model forecasts the length of each operating phase by drawing on existing hospital data. Estimates now land 50 per cent nearer actual pre-surgery time, 20 per cent nearer during surgery, and 40 per cent nearer post-surgery. Operating room planning is steadier as a result.
St. Luke's Medical Center — St Luke's Medical Center now runs Rapid AI, scanning CT and MRI scans for early stroke signs, including subtle warnings before a major episode. The hospital calls its AI stroke detection tool the first and only one in the Philippines, helping clinicians pick treatment faster through this AI-assisted neuroimaging approach.
Region Östergötland — Linköping University Hospital and Region Östergötland now run Stratipath Breast, a CE-IVD marked AI breast cancer diagnostics tool, in routine care. It delivers automated cancer risk profiling from standard pathology slides within 60 minutes, feeding Sectra's digital pathology platform already installed there.
The Nairobi Hospital — East Africa's first AI-powered CT scanner is now running at The Nairobi Hospital: a 256-slice Siemens SOMATOM Drive unit, unveiled by Health Cabinet Secretary Aden Duale for cardiology, oncology, neurology and trauma work. It scans faster, uses less radiation and produces sharper images, marking a leap forward in AI-driven medical imaging across the region. Alongside the launch, the Ministry of Health and the hospital agreed a partnership under the Social Health Authority to subsidise cancer and kidney treatment for patients unable to pay.
Ministry of Technology and Science (Zambia) — Twenty health posts across Zambia's 10 provinces are testing AI diagnostics in healthcare under a new pilot from the Ministry of Technology and Science, the Ministry of Health, 73Health Global and Pentatech Ltd. Finnish equipment allows clinicians to assess patients from afar, and the homegrown Larry RX system flags dangerous drug combinations before they reach a patient. Backers say pairing remote examination with automated prescribing checks should make treatment more uniform nationwide, though independent results are still pending.
Ministry of Health, Welfare and Health Infrastructure of Equatorial Guinea — Guinea Salud, Equatorial Guinea's first AI health chatbot, lets citizens text a WhatsApp assistant for free facts on illness, jabs, care and HIV. MCD Global Health gave technical help behind this AI-powered health assistant, made by the Ministry of Health, Welfare and Health Infrastructure. A trial stage covering knowledge topics runs now, before growth matched to public needs.
Addis Ababa City Administration — Chest scans at Yekatit 12 Hospital now pass through an AI-powered X-ray diagnosis tool built by RadiSen, deployed by Addis Ababa's city administration since 5 June 2025. Called AXIR-CX, the system performs AI chest X-ray analysis to spot tuberculosis, pneumonia and other lung abnormalities straight from digital images without waiting on a radiologist. Officials frame the rollout as part of Smart Living, a strand of the capital's Smart City agenda aimed at upgrading how healthcare gets delivered.
Vestre Viken Health Trust — Vestre Viken Health Trust has begun Norway's first trial of AI chest X-ray reporting, testing Oxipit's CXR Suite. ChestLink reads normal scans autonomously, skipping radiologist sign-off, while ChestEye screens for 75 lung pathologies as part of an AI-powered thoracic imaging check. Good results could push both tools out to other hospitals region-wide.
Hospital Universitario Docente Vladimir y Ilich Lenin — Hospital Universitario Docente Vladimir y Ilich Lenin in Holguín runs AI hospital administration for governance choices: tracking supply stocks, scheduling admissions and shifts, flagging vulnerable patients via dashboards that shorten waits. Cardiologist José Leandro Pérez Guerrero told ACN this mirrors AI adoption in Cuban hospitals.
Saudi Government and Seha Virtual Hospital — Seha Virtual Hospital and the Saudi Government screened hundreds of thousands of pilgrims with Lunit INSIGHT CXR at the 2025 Hajj, applying AI chest X-ray screening to catch tuberculosis, pneumonia and MERS. Between 4 and 9 June, the software processed huge scan volumes so staff could triage faster and treat sooner. This computer-aided radiology diagnostics approach replaced manual review with bulk automated analysis.
Region Uppsala — Region Uppsala now runs Stratipath Breast, a CE-IVD marked tool built for AI risk profiling for breast cancer, as part of everyday diagnostic routine. Ordinary pathology slides feed the software, which hands back prognostic output in under an hour, letting clinicians settle on treatment for early breast cancer sufferers with less delay. The approach works as an AI-based breast cancer prognostic tool that backs up, rather than substitutes, a pathologist's own judgement.
Rwanda Ministry of Health / Rwanda Biomedical Centre (RBC) — 58,567 community health workers across Rwanda now gain skills through a new digital platform from Rwanda Biomedical Centre and the Ministry of Health, replacing face-to-face sessions with mobile and online study. AI in community health training here picks tailored lessons, checks readiness and hands out certificates automatically. Linked to Rwanda's community electronic medical records system, the tool pulls live data to flag illness sooner, part of a push toward AI-driven health worker education.
Netcare — South Africa's Netcare now runs ICU deterioration prediction software across its full network of ICUs, a national first. The artificial neural network reads vital signs continuously and warns of sepsis, heart failure or respiratory compromise roughly ten hours ahead of visible symptoms, an AI early-warning system in intensive care built to give doctors more time to act. Clinicians can adjust medication earlier as a result.
Nawaloka Hospitals PLC — Nawaloka Hospitals runs NASHA, an AI-powered outpatient department chatbot reviewing patient symptoms and directing suitable care. A separate corporate chatbot sits on its website, both already active. Sri Lanka gets its first such setup for digital patient interaction, an early move into AI-driven OPD triage support.
Asklepios Kliniken — Germany's Asklepios hospital group is deploying an AI radiology platform, installing Aidoc's aiOS system at over 25 sites, already running at Hamburg-Altona, Hamburg-Wandsbek and Sylt. The software scans CT and X-ray images instantly, spotting stroke signs, brain bleeds, chest abnormalities and broken bones. Full coverage of every radiology unit with the AI diagnostic imaging tool is planned by the end of 2025.
Hospital María Auxiliadora — Sixty seconds or less: that is the new benchmark at Hospital María Auxiliadora in Lima, where an AI lung cancer detection tool now scans chest X-rays for signs of pulmonary neoplasia. Built through a partnership between Peru's Ministry of Health (MINSA), AstraZeneca and QURE.AI, the software drafts abnormality reports on its own, cutting the wait for early screening of lung and other pulmonary conditions. Clinicians gain a faster first read, powered by this AI-assisted chest X-ray screening approach.
Taichung Veterans General Hospital — Foxconn's Nurabot, built with Kawasaki Heavy Industries, is on trial at Taichung Veterans General Hospital as an AI healthcare robotics platform. It delivers medicine and specimens, patrols wards and educates patients — cutting nursing workload up to 30%, Foxconn's own unaudited figure. FoxBrain, Foxconn's large language model, and an NVIDIA Jetson AGX Orin chip supply the onboard AI compute, an early hospital automation case.
Sunway Medical Centre, Sunway City — Sunway Medical Centre (SMC) in Sunway City is the first hospital in Malaysia and Southeast Asia offering AI spine surgery, via the AiBLE Spine Surgical Suite. It merges live scans, 3D mapping and machine-guided spinal navigation so doctors can rehearse an operation on a virtual copy of the patient's backbone beforehand, a method used over 50 times already. It also covers cranial, orthopaedic, ENT and trauma work.
China-Guinea Friendship Hospital — Guinea's first AI-assisted hip replacement took place in Conakry, where China's 31st Chinese medical team, led by Diao Naicheng, joined local doctors at the China-Guinea Friendship Hospital for the AI-assisted total hip replacement. Software built a 3D map of the patient's bone anatomy, then weighed several surgical options to pick the best-fit prosthesis and implantation angle. The 90-minute procedure ran with specialists in China watching live over a video feed.
Institut de Cardiologie d'Abidjan / African Research Network — 300 hypertension patients in Côte d'Ivoire are testing an AI chatbot for hypertension patients called AVIA (Assistant vocal d'intervention automatisée), a voice-based AI health monitoring tool built by Ivorian startup data354 for the Institut de Cardiologie d'Abidjan and the African Research Network. Split into two groups, participants get calls every three weeks so it can track their condition, flag unusual readings to doctors and gauge blood pressure gains over six months. Only a phone line is needed, no internet, suiting patients far from clinics.
Charité – Universitätsmedizin Berlin — Germany's first hospital deployment of AI clinical documentation is underway at Berlin's Charité, where Microsoft's Dragon Copilot listens to doctor-patient exchanges and drafts notes automatically. Spoken consultations become structured records across several outpatient units and emergency rooms, under a six-month preview. Charité expects sharply reduced paperwork, freeing clinicians for direct patient contact.
Universitätsspital Zürich and Universitäts-Kinderspital Zürich — A pilot at University Hospital Zurich and University Children's Hospital Zurich uses an AI tumour board to sift medical records, scans and genetic profiles, flagging likely treatment paths for specialists. Humans still decide, with this AI-assisted oncology system acting purely as an advisory layer.
Ministry of Health Malaysia — Seven government clinics in Malaysia now run AI X-ray screening for lung disease among high-risk smokers and chemical workers. Machine-read scans sharpen diagnostic accuracy, says the Ministry of Health Malaysia, and nearby hospitals catch flagged cases for CT follow-up. Kit per site costs roughly RM10,000 upfront, per the ministry's figure.
ASL di Viterbo - Ospedale Santa Rosa — Santa Rosa hospital in Viterbo now spots cerebral stroke via AI stroke diagnosis software tuned for emergency speed. Tuscia network radiologists apply it to perfusion scans, marking ischaemic zones and cutting delays before treatment. Wired into the region's fibre-optic grid across local hospitals, this AI-powered stroke detection tool ran quietly for weeks beforehand.
Vorarlberger Landeskrankenhäuser (LKH Feldkirch) — Prostate cancer patients became the first in Austria to receive AI-assisted adaptive radiotherapy at LKH Feldkirch, delivered through the Elekta Evo linear accelerator's AI-assisted CT scanning capability. The machine reads scans moment to moment, reshaping each session's exposure around that day's body position and shielding surrounding tissue from unneeded dose. The launch establishes Austria's first use of AI-guided radiation therapy.
Hospital de Bellvitge — AIINANE, an AI preoperative risk assessment tool built entirely by Hospital de Bellvitge's own anaesthesiology team, is now in clinical trial. It mines electronic health records plus a patient-completed digital questionnaire, flagging complex cases for clinician attention while sparing lower-risk patients extra pre-surgery appointments. Testing began with breast cancer day-surgery patients, with wider use across ambulatory procedures targeted for mid-2026 — part of a broader push toward AI-driven surgical risk prediction.
Tsinghua University — China now has an AI hospital run entirely by artificial intelligence: Tsinghua University's AI Agent Hospital fields 42 AI doctors spanning 21 departments and diagnosing over 300 conditions. An internal trial platform, active since November 2024, reached 93.06% accuracy on respiratory illness after processing roughly ten thousand simulated patients, giving clinicians a working AI clinical diagnosis benchmark. Beijing Tsinghua Changgung Hospital will host the initial rollout.
Grupo NIC (Vitalics) — Grupo NIC's Vitalics gives Paraguay its first automated AI triage booth, scanning blood pressure, temperature and heart rate before an AI-guided questionnaire flags next steps and drafts a preliminary report. Running 24/7, it targets shorter hospital queues and wider primary healthcare reach via this self-service AI health kiosk.
Spitalul Clinic Județean de Urgență Bihor — Bihor County Emergency Hospital in Oradea runs Romania's first AI in healthcare deployment via ZenAiOS, a hospital AI platform for medical units. Modules serve patients, doctors, nurses and managers; 1,800-plus patients have used Zena, the patient-facing tool. Doctors get a digital assistant trained on hundreds of medical textbooks.
Helse Nord-Trøndelag (HNT) — Sykehuset Levanger, under Helse Nord-Trøndelag, has become the launch site for AI fracture detection radiology through BoneView, an algorithm built by Gleamer, kicking off Helse Midt-Norge's Mid-Norway trial. Within one to two minutes it spots breaks, joint fluid and dislocations, meaning patients with clear results can go home before an orthopaedic check happens, though radiologists still examine every picture. That AI skeletal X-ray analysis forms part of a two-year study across all the region's health enterprises.
Hospital Universitario Nuestra Señora del Rosario — Spain's first AI-based coronary diagnosis centre is Madrid's Hospital Universitario Nuestra Señora del Rosario, now using DeepVesselFFR, built with Kanbai Tech, to read CT scan data via 3D-modelling and gauge coronary blood flow without catheterisation. It cuts unnecessary invasive procedures by up to 70%, delivering results in six hours. This non-invasive coronary artery evaluation relies on a developer-reported figure, not independent audit.
MAPFRE Guatemala — MAPFRE Guatemala relaunched its AI telemedicine platform MedicOnline on 2 April 2025: a smartphone scan of vital signs precedes each remote doctor's visit. This AI-based vital sign scanning flags heart rate, blood pressure, cholesterol risk, and type 2 diabetes risk beforehand, free and available 24/7 for MAPFRE's insured.
Akershus University Hospital (Ahus) — AI fracture detection is live at Akershus University Hospital's Nordbyhagen site: Gleamer's BoneView reads X-rays, after Ski and Kongsvinger. Norway's busiest orthopaedic unit runs 30,000 X-ray scans yearly through the AI X-ray diagnosis tool, flagging breaks in under two minutes. Estimated: 8,000 freed consultations yearly, 15,000-20,000 fewer waiting hours — unaudited trust figures.
MedLife — MedLife has added an AI assistant to its mobile app that reviews two years of clinical results and six months of lab data to build a personalised health summary, then suggests which medical specialty a patient should book a follow-up with. As Romania's largest private healthcare provider, MedLife frames the tool as part of a wider shift toward AI health guidance, drawing on accumulated patient history rather than a single visit to shape its suggestion. This AI-powered personalised health summary depends on the accuracy and completeness of records already stored in the app, so its recommendations are only as dependable as the underlying data.
Capio — Capio, a major private healthcare provider in Sweden, is expanding AI-driven medical documentation across all business areas after successful pilots. Ten clinics in primary, specialist and hospital care reported lighter workloads and more consultation time. Tandem drafts clinical notes automatically from live conversations, a method some call AI clinical note-taking.
Guang'anmen Hospital — A TCM AI diagnosis model, Guangyi-Qizhi, has gone live at Guang'anmen Hospital in Beijing, built jointly with Baidu Intelligent Cloud and Quanzhen Medical. It reads multimodal signals such as tongue and pulse data to draft records in 2-3 seconds, then guides physicians through syndrome differentiation, prescriptions and medication choices across the whole path from intake to discharge follow-up, serving as an AI-assisted traditional Chinese medicine documentation system.
SURA — At Colombia's Centro de Ciencias Ómicas, Sura applies generative AI in genomic medicine, running Microsoft technology across genetic records from more than 10 thousand patients. Staff now type plain-language questions into a Teams interface for AI genomic data analysis instead of sifting through spreadsheets by hand. This speeds up treatment planning and personalises care across the insurer's five million-strong membership.
Discovery Health — South African insurer Discovery Health built Personal Health Pathways, applying AI in healthcare personalisation via machine learning and digital twin modelling to tailor treatment advice for adult medical aid members. Wearable data, claims records and health files train this personalised medicine platform to spot fitting interventions. Discovery frames the payoff as better outcomes, lower cost — an unaudited, company-stated benefit.
West China Hospital of Sichuan University — Sichuan University's West China Hospital has switched on Huaxi Hongyi, a 72-billion-parameter medical large language model now running across hospital departments for AI in healthcare applications, unveiled on 19 March 2025. Engineers fused over 10 general-purpose systems with more than 50 specialist ones, feeding the build with 1.1 million doctor-patient dialogue records, 1,600 clinical guidelines and 3.5PB of scanned imaging data. Among the clinical large language model's live jobs: sorting incoming patients, drafting medical records automatically, spotting early-stage cancers during endoscopy, backing up staff on shock-response calls, and staffing a health-outreach line that has logged over 1.2 million calls.
Centre Hospitalier Princesse Grace — About 30% more MRIs now pass through Centre Hospitalier Princesse Grace (CHPG), up from around 6,000 yearly, after its scanner gained AI-assisted MRI through the SmartSpeed AI module. Within three months, over 1,500 patients had already used the AI-powered MRI scanner, Dr Mathieu Liberatore noted, pointing to sharper, quicker imaging behind faster diagnoses. A gift from Dmitry Rybolovlev covered the cost.
Hospital Universitario de Caracas — Hospital Universitario de Caracas now runs an AI-powered MRI machine, called by Delcy Rodríguez the nation's first scanner with embedded AI. Radiologists keep final say; the system only flags findings. Officials call it part of a wider shift toward AI-assisted imaging in Venezuelan hospitals.
IRCCS Istituto Clinico Humanitas — Every Humanitas Medical Care site in Lombardy, Piedmont and Sicily now offers AI breast cancer screening, after Humanitas widened a system first trialled in Milan. The tool reviews mammogram quality instantly, calculates breast density and flags likely lesions for radiologists during early detection of breast cancer — a first for an Italian hospital group.
Cliniques Universitaires de Kinshasa — Kinshasa's Cliniques Universitaires began an AI stethoscope diagnosis trial of the Pneumoscope, checking whether it detects pneumonia, bronchopneumonia, bronchiolitis and asthma, with sites in Beni. Researchers compare its readings against clinical exams and imaging for accuracy, a form of machine-assisted auscultation for respiratory disease.
Sindh Government Qatar Hospital — Sindh Government Qatar Hospital in Karachi's Orangi Town now runs the province's first public AI X-ray machine for TB diagnosis, the Rs45 million Cat 4 Integrated system. In five minutes it converts grayscale scans into colour-coded pictures, marking diseased spots in red to catch TB and chest infections, a tuberculosis screening tool built for speed. Around 120 to 150 sufferers visit daily.
Health New Zealand — Health New Zealand Te Whatu Ora is trialling AI diabetes retinal screening in South Auckland, aimed at the Pacific community. Seven general practices deploy trained personnel to capture eye photographs; software flags concerns, while an ophthalmologist verifies outcomes. Roughly 840 people should be seen across three months, with AI-based eye screening for diabetes validated on 3,000-plus patients should the concept prove successful.
Hospital Materno Infantil — Hospital Materno Infantil in Tegucigalpa bought an AI-powered CT scanner for 17 million lempiras, among the most advanced AI imaging machines in Honduras's public health system. The Aquillon Serve unit's built-in AI flags lung nodules, runs brain spectral scans, and digitally subtracts images for cerebral circulation studies. It also helps identify cancerous tumours and measure their size.
Fujian University of TCM Affiliated People's Hospital — First in Fujian province: this hospital now runs DeepSeek R1 70B locally, linked into record-keeping and screening systems. The Affiliated People's Hospital of Fujian University of Traditional Chinese Medicine gives doctors an AI hospital deployment for analysing patient data and drafting diagnoses fast. Next, a TCM-specific decision support agent extends this clinical AI integration into traditional Chinese medicine.
AI4PEP Senegal — AI4PEP Senegal built Plateforme 3S One Health, an AI-powered zoonotic disease surveillance platform for tracking outbreaks at community level. Three Wolof tools drive it: Taggàt, a chatbot; Jottali, for alerts; and Gëstu, a risk dashboard — linking automated public health surveillance to epidemic detection.
Hospital Authority — Six Hong Kong hospitals will pilot generative AI medical reports under a Hospital Authority scheme unveiled 20 February 2025: Princess Margaret, Caritas Medical Centre, United Christian, Tseung Kwan O, Queen Elizabeth and Tin Shui Wai. Chief Executive Tony Ko Pat-sing called it a self-built AI-assisted clinical documentation tool, aimed at letting clinicians pull up case files faster and generate first-draft write-ups automatically. Staff physicians must still confirm and sign off on every output before it becomes official.
Hospital Militar Comandante Manuel Piti Fajardo, Hospital Provincial Arnaldo Milián Castro, Hospital Oncológico Dr. Celestino Hernández Robau — Three Villa Clara hospitals in Cuba pilot AI medical imaging from Universidad Central Marta Abreu de Las Villas. Hospital Militar Comandante Manuel Piti Fajardo reads X-rays for COVID-19, Hospital Provincial Arnaldo Milián Castro flags lung nodules, and Hospital Oncológico Dr. Celestino Hernández Robau screens breast masses. As a second reader, these AI diagnostic imaging systems surface patterns a radiologist might miss.
Acıbadem Sağlık Grubu — Acıbadem Health Group has built AI imaging in healthcare into its newly opened Istanbul site, Acıbadem Kartal Hospital, fitting deep learning CT, 3 Tesla MRI and 3D mammography under one roof. Doctors get sharp diagnostic output from just one pass, while fetal abnormalities surface quickly through machine-assisted scan review.
Radboudumc, Ziekenhuis Rivierenland, Gelre ziekenhuizen, Ziekenhuisgroep Twente, Catharina Ziekenhuis — Radboudumc, Ziekenhuis Rivierenland, Gelre ziekenhuizen, Ziekenhuisgroep Twente and Catharina Ziekenhuis now share three diagnostic algorithms through the national Twiin Portal under the Dutch AI For Imaging (AIFI) pilot. Radiologists using this radiology AI in hospital diagnostics gain a second check on fracture detection, incidental pulmonary embolism spotting and skeletal age estimates, one of the earliest coordinated deployments of medical imaging AI across several hospitals at once.
Peking Union Medical College Hospital — PUMCH-Taichu, a rare disease AI diagnosis model jointly built by Peking Union Medical College Hospital and the Chinese Academy of Sciences' Institute of Automation, entered clinical use in China. Drawing on years of accumulated rare disease knowledge and genomic data specific to the Chinese population, and combining DeepSeek-R1 reasoning capabilities, the system is the first large model designed for this group. It supports faster, more precise identification of rare conditions and also lets patients hold initial consultations through multi-turn dialogue, offering AI-assisted rare disease identification directly to the public.
Medicover Romania — Medicover Romania applies AI in medical imaging at five sites — Bucharest, Cluj-Napoca, Constanța, Craiova, Iași — using GE HealthCare's AIR Recon DL and CT auto-positioning. MRI and CT scan times fall about 30%, with sharper accuracy, an AI-based diagnostic imaging rollout at Romania's largest private healthcare operator.
Cedars-Sinai Medical Center — Cedars-Sinai Medical Center is trialling AI in nursing documentation via Aiva Nurse Assistant, letting staff speak observations that flow straight into Epic once a clinician signs off. A 48-bed surgical floor hosts the trial ahead of a wider inpatient rollout. This mobile app for nurses swaps manual charting for spoken input.
Department of Health and Social Care — 700,000 UK women will join EDITH (Early Detection using Information Technology in Health), a £11 million Department of Health and Social Care trial of AI breast cancer screening spanning 30 NHS sites. One radiologist using the AI mammogram reading tool could match output now needing two, a change that might release hundreds of specialists elsewhere.
Meuhedet Health Fund — AI in healthcare lands at Meuhedet Health Fund, where Microsoft Copilot now spans clinical work and administration. Doctor Copilot listens to consultations, writes up notes and proposes referrals; Copilot 365 lifts general office output. Phone and digital queries from patients trigger automated clinical documentation replies, plus adapted options for the Haredi community.
Hospital Nacional Daniel Alcides Carrión — Callao's Hospital Nacional Daniel Alcides Carrión now runs an AI-powered MRI scanner, its first, plus a new AI mammography unit, both part of a medical imaging AI expansion. Tumours get caught at clinical stages 1 and 2, where survival odds top 90%, and the scanner speeds up readings for neurology, oncology, cardiology and orthopaedics. Priced above 21 million soles, it should cover 19,710 patients yearly.
AOU San Luigi Gonzaga di Orbassano — San Luigi Gonzaga hospital in Orbassano now runs AI avatar surgery guidance inside the operating room, steering robotic prostate and kidney cancer procedures by comparing each patient's clinical profile against data from hundreds of prior cases. Built by the hospital's urology unit with the University of Turin and the University of Calabria, the system answers surgeons' spoken questions and adjusts its guidance to the individual being treated, an early real-world case of conversational AI surgical assistance in live procedures.
South Australia Medical Imaging (SAMI) — South Australia Medical Imaging (SAMI) has become Australia's first public radiology provider offering AI radiology diagnosis support across an entire state network, adopting Annalise.ai's chest X-ray tool at ten hospitals. Among SAMI's 1,200-member workforce, this AI diagnostic support flags suspicious regions on scans and proposes likely diagnoses, though clinicians still confirm every outcome.
Ministry of Health Brunei Darussalam — BruHealth 5.0's new Smarter Health Index lets Brunei's Ministry of Health apply AI in healthcare to diet, sleep and stress data from 16 January 2025, generating tailored wellbeing steps. Built by EVYD Technology, it adds cervical cancer to existing breast, colorectal and cardiovascular disease screening. AI-generated health guidance now anchors the country's public health platform.
Belau National Hospital — VeriSee DR, AI software screening for diabetic retinopathy, now runs at Palau's Belau National Hospital via a joint venture with Taiwan's Shin Kong Hospital. Their Taiwan-Palau Smart Hospital Project adds remote pathology review and secure cross-border imaging transfer. Pacific hospitals increasingly use AI-assisted diabetic eye screening where specialist access is scarce.
ENEA — AI drones penguin counting near Mario Zucchelli Station in Italy's Antarctic territory now reaches 97% accuracy for grown Adélie penguins and 89% for chicks, according to the Italian national agency ENEA. Researchers there paired aerial cameras, a trained neural network and the CRESCO supercomputer, cutting reliance on manual headcounts by field crews. Drawing on over 400 images gathered across eight earlier expeditions, the software tells adult birds apart from chicks in hard-to-reach colonies, giving wildlife population monitoring by AI a quicker, sharper edge over counting by eye.
Universitätsklinikum Schleswig-Holstein (UKSH) — MAIA (Medical Artificial Intelligence Assistant) is Germany's first hospital-based clinical decision support system, now live at Universitätsklinikum Schleswig-Holstein. Built with Tiplu GmbH and certified as a Class IIa medical device, it flags clinicians to danger signs of kidney failure, sepsis or falls in real time. Kiel's internal medicine unit adopted it first, ahead of expansion into intensive care and emergency units at Kiel and Lübeck, with full inpatient coverage of this medical AI assistant planned.
Fundación Cardioinfantil (LaCardio) — Colombia's first clinical AI unit, marking a new stage in AI in cardiology, has opened at Bogotá's LaCardio (Fundación Cardioinfantil): the Unidad Clínica de Inteligencia Artificial (UCIA). Built on digital twin organ simulations, a Framingham cardiovascular risk score and mortality-prediction models, it aims at sharper heart disease detection and treatment through data-driven clinical decisions.
Brainomix 360 e-Stroke, CT-analysis AI from an Oxford University spinout, guides treatment and transfer decisions across NHS stroke networks in England. A Royal Berkshire Hospital study in Frontiers in Neurology linked it to a tripling in patients reaching functional independence, 16% to 48%, and a real-world evaluation found mechanical thrombectomy up over 50%. The study was pre/post observational, not randomised, and the vendor-reported scale of about 60,000 patients at every English stroke centre is unverified.
Quebec Ministry of Health — Vitr.ai's AI-powered patient triage tool Navig has replaced nurse- and physician-led first assessment across all 766 GMF (Groupes de médecine de famille) clinics in Quebec, from 1 April 2025, under a $40 million contract. This AI triage software interviews arriving patients through a secretary, then assigns each to a pharmacist, psychologist, physiotherapist or doctor as appropriate.
Ministry of Health (MINSA) — AI in hospital care now runs at 14 central hospitals across Angola, covering diagnostic imaging, clinical analysis, surgery and personalised medical training, the Ministry of Health confirms. Leonardo Inocêncio, Secretary of State for the Hospital Area, links this to fewer clinical errors and sharper accuracy from AI-assisted diagnostic imaging. Some 38 thousand health workers are gaining digital skills to run the new equipment.
Wiener Gesundheitsverbund (WIGEV) — AI-enhanced MRI is now live at two Vienna hospitals: Wiener Gesundheitsverbund has fitted Kliniken Donaustadt and Floridsdorf with deep learning reconstruction, trimming acquisition time by as much as 73% and turning a 15-minute knee scan into just 8. Klinik Favoriten and AKH Wien come next.
Klinikum Wels-Grieskirchen — Cardiac AI diagnostics arrived at Klinikum Wels-Grieskirchen via new Quantitative Flow Ratio (QFR) software, among Austria's earliest such catheter-lab rollouts. Angiogram images become 3D reconstructions feeding haemodynamic models gauging whether a stenosis meaningfully blocks blood flow, an AI-powered heart diagnostics step sparing patients invasive pressure-wire checks.
Ordensklinikum Linz — Roughly 50% fewer falls now occur on the acute geriatrics ward at Ordensklinikum Linz Barmherzige Schwestern using AI fall detection in hospital wards built on radar, not cameras. Overhead sensors track patients continuously; an AI module scores danger and alerts nurses' phones the instant it senses odd movement — a camera-free patient monitoring method behind the drop.
Gold Coast Health — First in the Southern Hemisphere: Gold Coast University Hospital now runs CAS-One IR, an AI-guided tumour ablation system reaching tricky liver and kidney lesions. Radiologists gain live 3D imaging plus AI-assisted comparison of pre- and post-treatment CT scans, confirming success on the spot rather than after follow-up. Deployment began July 2025.
Ramsay Health Care — Speech-to-text software now drafts clinical notes at St Andrew's Ipswich Private Hospital in Queensland, under a Ramsay Health Care trial of AI clinical documentation branded Ramsay Scribe. Built with T-Pro, it targets less paperwork, faster records — the firm's first deployment of this AI-powered medical notetaking tool.
Botswana national TB programme — Botswana's national TB programme has deployed Qure.ai's CAD system for AI chest X-ray screening for TB, aiming to speed up tuberculosis diagnosis. The tool flags likely TB cases from X-ray images for clinician review, and local staff have completed training on the AI-assisted tuberculosis diagnosis system.
Beijing Children's Hospital — Beijing Children's Hospital in China paired an AI paediatrician with 13 specialists to assess an eight-year-old boy. Baichuan Intelligence and Xiao'erfang Health Technology trained it on records from over 300 paediatric experts before a month-long internal trial preceded this AI in paediatrics rollout. Diagnostic accuracy beyond that trial remains unconfirmed.
Audere — Burundi's Cibitoke and Kirundo health districts began running Audere's Malariya Pi in August 2025, an AI malaria diagnostic tool built on the firm's HealthPulse AI engine. Frontline staff — from village-level workers through clinic personnel to lab technicians — rely on it to interpret rapid diagnostic tests for malaria, log patient histories and treatment progress, then escalate difficult cases upward to hospital care. The name itself means "bye bye malaria" in Kirundi, this AI-powered malaria case management app's semantic nod to its public health mission.
Al Hamshari Hospital — Rhazes AI, a UK-Qatar startup, is running AI clinical scribes at Al Hamshari Hospital near Ein el-Hilweh refugee camp in southern Lebanon, a Palestine Red Crescent Society site. Live transcripts, diagnostic prompts and finished structured notes now lighten workloads where staff face 60 patients daily. A trial lasting through November 2025 checks whether the AI medical documentation tool speeds paperwork and boosts physician confidence and patient flow.
Caja Costarricense de Seguro Social (CCSS) — A million-plus health records mined by an algorithm at Costa Rica's Clorito Picado Clinic in Tibás are the basis for a new AI diabetes risk prediction pilot run by the Costa Rican Social Security Fund (Caja Costarricense de Seguro Social, CCSS). Drawing on the Single Digital Health Record (EDUS) database, the tool marked 2% of patients as likely candidates for type 2 diabetes, and follow-up exams verified 132 had already progressed to the disease since 2023 - those individuals now receive ongoing monitoring to halt further deterioration. Early findings surfaced within months, paving the way for broader predictive healthcare analytics targeting acute coronary syndromes, retinopathy and cancer.
Fimlab Laboratories — Breast, prostate and gastrointestinal tissue samples at Fimlab now get a first read from AI, not a human. Aiforia and Proscia supply the platforms behind this AI-assisted pathology diagnostics rollout across the Finnish lab network, scanning digitised slides before a pathologist signs off on the final call. Behind the move: caseloads climbing faster than the pool of trained specialists, pushing labs toward AI in pathology to close the gap.
Fondation Lenval — Fondation Lenval, the first paediatric university hospital in Provence-Alpes-Côte d'Azur, runs a Siemens Somatom Go Top scanner in Nice from 1 September 2025 for AI-assisted medical imaging. The 128-slice machine personalises scan settings per child, aligns them via 3D camera, and helps read results, including spotting lung nodules — cutting radiation exposure by 30% without losing image clarity, reflecting wider hospital use of AI-powered radiology.
Klinikum Landsberg am Lech — Documentation workloads at Klinikum Landsberg am Lech, a 218-bed acute hospital in Bavaria, fell by 90 per cent once staff began using AI medical transcription running on Microsoft Azure OpenAI, sparing clinicians up to two hours each day. The speech-to-text clinical documentation tool has since processed more than 1,300 patient conversations, turning almost one million spoken words into text. Archiving and transcription expenses dropped 25 per cent, a figure the hospital itself reports rather than an independent audit.
Agaplesion Elisabethenstift Darmstadt — Colour-coded boxes flag suspected fractures under AZmed, an AI fracture detection tool running at Agaplesion Elisabethenstift in Darmstadt. The system also screens X-rays for tissue fluid and pneumothorax, backing up emergency-room staff when no radiologist is on-site, chiefly nights and weekends. Wider rollout across Agaplesion's hospital network is under consideration as an AI-assisted X-ray analysis approach.
Mediterraneo Hospital — Greece's first hub-and-spoke stroke network now runs on AI stroke diagnosis from Nicolab's StrokeViewer platform at Mediterraneo Hospital in Athens, linking the capital to island sites on Santorini and Paros. Real-time scan analysis lets remote doctors judge severity and choose on-site care or urgent thrombectomy transfer. This AI-driven stroke care reaches islands that previously lacked rapid specialist assessment.
Khoula Hospital — Muscat's Khoula Hospital now runs an AI spinal MRI scans platform, built with technology partner Civilization, that drafts preliminary diagnostic reports directly from imaging data. Clinicians get a head start on decisions instead of writing findings from a blank page. The rollout sits within a broader AI-powered radiology platform pilot programme overseen by Oman's Ministry of Transport, Communications and Information Technology.
Almoosa Health Group — A tablet-based AI physician named Dr Hua now greets patients at what its backers call the world's first AI doctor clinic, opened by Almoosa Health Group with Shanghai firm Synyi AI in Al-Ahsa, Saudi Arabia. It scans images and drafts treatment plans, though a clinician must still sign off before anything proceeds. Synyi AI puts its failure rate below 0.3%, a company claim not independently checked, and for now only respiratory cases fall under this AI-assisted diagnosis effort.
University of Aleppo — Tebyan now runs live at Aleppo University Hospital, the first clinical AI system inside a Syrian public university hospital. University of Aleppo Faculty of Medicine students and graduates coded it themselves, turning lab test results into structured PDF summaries and giving ward teams an AI clinical assistant for faster diagnoses.
SEHA (Abu Dhabi Health Services Company) — SEHA's AI-assisted tele-stroke robot, LEO360, spared 80 per cent of patients an unneeded transfer between Sheikh Tahnoon bin Mohammed Medical City and Al Tawam Hospital. Distant specialists reach a diagnosis via video, EMR data and automated guidance, needing just 10.7 minutes on average. This remote stroke care technology now underpins urgent care at the UAE's largest hospital operator.
Ministry of Health and Family Welfare — India's health ministry runs MadhuNetrAI for diabetic retinopathy screening at 38 sites in 11 states, built with AIIMS and Wadhwani AI. Staff read retinal images unaided, part of this diabetic eye examination AI push. 14,000+ images processed, 7,100 patients reached.
Primaya Hospital Group — AI hospital chatbot Sasya covers patient queries across all 20 Primaya Hospital Group hospitals in Indonesia, handling appointments, insurance checks and doctor selection. Running on Google infrastructure, this healthcare-bot deployment generated billions of rupiah by October 2025, with referrals driving roughly 30% of web bookings.
CHU de Guyane — Cayenne's public hospital, CHU de Guyane, now screens for bowel cancer using Cad Eye, an AI-assisted colonoscopy tool built into a high-resolution endoscope. Adenoma detection has jumped from 15 to 40 percent since the technology entered its gastroenterology department, which handles over 1,200 procedures a year. French Guiana thus joins the growing list of regions applying AI-powered colorectal cancer screening in routine practice.
Queen Ingrid's Hospital — Queen Ingrid's Hospital in Nuuk installed an AI-powered MRI scanner from GE HealthCare for sharper, quicker scans. This brings AI-assisted MRI scanning to Greenland's stretched healthcare system, letting it handle more patients despite thin local infrastructure. Rigshospitalet in Copenhagen can now run it remotely, sparing patients the 3,500-kilometre trip to Denmark.
Ghana Health Service — Ghana Health Service now runs AI TB screening at 37 Military Hospital in Accra using a Qure.ai X-ray system. It flags pulmonary abnormalities within seconds, easing a shortage of radiologists and catching undiagnosed cases via the Active Case Finding project. The AI-powered X-ray diagnosis tool targets faster, more accurate results.
Catharina Ziekenhuis, UMC Utrecht — Faint lung nodules once easy to overlook are now caught earlier at Catharina Ziekenhuis in Eindhoven and University Medical Center Utrecht, thanks to Qure.ai's qXR software running on ordinary chest X-rays, an application of AI in radiology for lung cancer detection. The CE-certified system highlights suspicious spots no matter why a scan was requested, so people X-rayed for unrelated reasons still gain a benefit. Both hospitals set up new opportunistic screening pathways around this incidental lung nodule detection on chest X-rays, turning routine imaging into an added chance to catch cancer sooner.
Catharina Ziekenhuis — Catharina Ziekenhuis in Eindhoven now runs Pacmed Critical, delivering AI ICU discharge decisions to intensivists on the Intensive Care Unit. The hospital becomes the third within the Santeon network to adopt this software, and the first to integrate it directly with the HiX electronic patient record system. Health insurers Zilveren Kruis and CZ back the wider project, which extends AI decision support in intensive care to further hospitals.
Máxima MC — Since 2025, radiologists at Máxima MC have relied on AI fracture and haemorrhage detection to spot X-ray fractures alongside CT findings of pulmonary embolisms, cervical spine breaks and brain bleeds. This diagnostic support tool complements clinical judgement; staff retain final responsibility. Broader rollout across specialities is under review.
Instituto Nacional de Enfermedades Neoplásicas (INEN) — AI MRI cancer diagnosis at Peru's National Institute of Neoplastic Diseases (INEN) has cut 1.5 Tesla scan times from 50 to 30-40 minutes. Embedded software adjusts for patient motion automatically, removing the need to stay still. This AI-assisted MRI scanning setup now handles roughly 1,000 studies monthly.
Centro Hospitalar Universitário de Lisboa Central (CHULC) / Hospital Curry Cabral — Curry Cabral Hospital in Lisboa became Portugal's first public hospital to combine AI-assisted endoscopy with robotics (IA) for digestive tract procedures. The setup pairs automated lesion-detection algorithms with a robotic assist arm for greater precision during examinations, adding robotics in digestive endoscopy to routine practice. Installation took place at the start of October 2025.
Hospital Particular da Madeira — Hospital Particular da Madeira is now the first facility in the Região Autónoma da Madeira using AI-assisted radiology in its emergency department. Gleamer's software, with BoneView and ChestView modules, applies machine-learning image analysis to chest and bone X-rays, flagging fractures, nodules and other lesions in Portugal.
Ústredná vojenská nemocnica SNP Ružomberok – Fakultná nemocnica — Ružomberok's military hospital now runs AI radiotherapy planning through Limbus Contour AI, shrinking organ-at-risk contouring to two minutes. High-risk prostate cancer cases see total planning time cut by 26 minutes, since the software delineates anatomical structures automatically. This marks one of Slovakia's first uses of AI-assisted contouring in radiation oncology.
Penta Hospitals SK — Slovakia gained its first nationwide hospital use of AI X-ray diagnostics as Penta Hospitals SK activated the technology at all 19 sites with radiology units. Gleamer and Rayscape built the tools, spotting pathological changes on chest and skeletal scans, marked on-screen. The AI-assisted radiology rollout began autumn 2025.
National University Hospital (NUH) — MedBot, National University Hospital's AI medication counselling assistant, cuts 28 staff-hours monthly and about S$15,400 yearly since July 2025 across 66 approved drugs. It greets patients before prescription pickup at kiosks in NUH's Kent Ridge Wing Pharmacy and National University Centre for Oral Health pharmacy. Pharmacists review each reply from this generative AI pharmacy assistant first.
Ministry of Health of Southwest State — A 2025 partnership between Somalia's Southwest State health ministry, ThinkMD and Cleanport Ventures is putting AI clinical decision support into the hands of frontline caregivers. The ThinkMD app steers phone-based checkups through a structured, WHO-aligned sequence and surfaces science-backed treatment guidance as patients are seen. Authorities describe it as the nation's inaugural state-run AI-supported patient assessment scheme.
Médecins Sans Frontières (MSF) — MSF is testing AI snake identification at two South Sudan hospitals, Twic and Abyei, to speed bite treatment decisions. Trained on 380,000 snake photographs, the machine-learning system predicts venomous versus harmless species, guiding correct antivenom use. The early-stage trial shows encouraging, non-conclusive results.
Hospital del Mar, Barcelona — Barcelona's Hospital del Mar has introduced an AI voice agent for COPD monitoring after patients go home, using LOLA, Tucuvi's clinically validated system, to phone them daily and catch trouble early during the critical first 30 days post-discharge. Rather than waiting for a crisis, LOLA tracks whether patients keep up with prescribed treatment, watches for red flags and pings clinicians the moment an exacerbation looks likely, following profiles that pneumology staff built for the tool. The aim is simple: keep people out of hospital beds through consistent remote patient monitoring.
Klippan Municipality — Klippan municipality trialled AI documentation for social workers using software called Tandem, with participants gaining an extra hour daily previously spent on paperwork. Among 16 people involved, 94 per cent rated the AI-assisted documentation for care staff more favourably than older methods, and 69 per cent felt fully engaged during client sessions across the one-to-three-month trial.
University of Dodoma (UDOM) — University of Dodoma (UDOM) built Akili Check, an AI mental health platform live at Muhimbili National Hospital, Benjamin Mkapa Hospital, Mirembe Psychiatric Hospital and Amana Hospital in Tanzania. Over 200 patients have been diagnosed and referred to care through it. The AI-driven psychiatric screening tool combines self-assessments with treatment guidance.
Ministry of Health of Ukraine — Ukraine's Ministry of Health screened over 870 diabetes patients in Odesa and Chernihiv for AI-powered eye screening for diabetic retinopathy via CheckEye, October 2024-March 2025. Scans took under 30 seconds; 25% needed ophthalmologic follow-up, 75% did not. The pilot used automated retinopathy detection as triage, not diagnosis.
Zaporizhzhia Regional Clinical Hospital — A CheckEye AI eye screening trial at Zaporizhzhia Regional Clinical Hospital flagged possible sight-threatening complications in one of every ten outpatients. Over two months, more than 9% showed diabetic retinopathy of differing severity, several undiagnosed diabetics. This suggests retinal AI screening can surface disease standard records miss.
Alliance for Public Health — Over 7,000 people use TWIIN, the Alliance for Public Health's AI digital health assistant from the IAS 2025 conference. Reachable 24/7 via websites, apps and API links, it fields HIV, STI, mental-health and harm-reduction questions by spoken or written exchange. That figure is self-reported, unverified.
Ministry of Public Health — A 25-hospital trial of Thailand's new AI drug interaction alert system checked nearly 90,000 medication records, catching 2,384 risky combinations and firing over 3,800 live warnings that pushed doctors to revise care before dispensing. Thailand's Ministry of Public Health now runs the tool in 11 hospitals, targeting every government hospital by September 2025 — a broad national push for AI-driven prescription safety screening.
UNDP Turkmenistan — AI-powered tuberculosis screening is now running nationwide in Turkmenistan, rolled out by the United Nations Development Programme (UNDP) with Global Fund funding. CAD4TB software reads X-rays in seconds at central and velayat hospitals, and each portable unit manages roughly 400 exams per charge on solar or battery power. This tuberculosis detection using AI reaches remote regions beyond fixed clinics.
Urgent Care Medical Centre — Urgent Care Medical Centre is trialling AI chest pain diagnosis and forecasting software from Scopium AI. Early-stage checks started at St George, with full clinical evaluation in July. Davidson Ishmael, Minister of State in the Ministry of Health and Wellness, views the AI cardiac screening as a Caribbean first for Barbados.
FLENI — Buenos Aires neurology hospital FLENI is piloting Navian, bringing AI in brain surgery to about 20 procedures, projecting augmented-reality 3D reconstructions of tumours, veins and nerves onto patients. Three Argentine ITBA bioengineering graduates built the AI-assisted neurosurgery system for ordinary smartphones or AR glasses, costing a fraction of neuronavigators, which can top US$1 million.
Hospital General de Agudos Parmenio Piñero — Hospital Piñero's geriatrics unit in Buenos Aires is piloting AVIA (Algoritmo de Validación de Independencia y Autonomía), an AI predicting loss of autonomy in older patients with 75% sensitivity, across 200 patients. Fundación Cuidados de Salud built this predictive model for functional decline in ageing populations on the British ELSA database's 20 years of data, enabling early intervention.
Island Health — Joint scans at Nanaimo Regional General Hospital now take five minutes, down from 10-15, using trial AI-enhanced MRI. Island Health made the AI-assisted MRI software permanent, expanding it to Royal Jubilee Hospital, Victoria General Hospital, Campbell River and Comox Valley.
Osaka International Cancer Center, NIBN and IBM Japan — Two generative AI in healthcare systems are now running clinically at Osaka International Cancer Center, built with Japan's NIBN and IBM Japan: one turns paper intake questionnaires into electronic-record entries, the other drafts documentation from nursing handovers and phone calls. Under this AI-driven patient care approach, symptom-hearing time is expected to fall to 25% of what it previously took, and testing found the transcription tool trimming charting effort by around 40%. Neither figure has been checked by an outside party — both come directly from the developers.
Hospital Ángeles — Nine minutes: that's how fast AISHA MRI turns a magnetic resonance scan into a full-body 3D digital twin, spotting over 290 anatomical structures with better than 90% accuracy. Rolled out across all 27 Mexico sites run by Hospital Ángeles, this AI medical imaging diagnostics tool has revealed that a third of patients carry a significant finding warranting further review — an early proof point for AI-driven medical imaging at scale.
Ministry of Health and Social Services — Namibia's Ministry of Health and Social Services deploys AI-powered digital chest X-rays for AI tuberculosis screening and earlier diagnosis. On World TB Day in the Omusati region, Minister Esperance Luvindao pledged support for extending AI-aided digital X-ray screening to health facilities and communities. Its TB incidence ranks among the world's 11 highest.
SPMS (Serviços Partilhados do Ministério da Saúde) — SPMS built an AI respiratory symptom triage tool into Portugal's SNS 24 phone and app service, launched under the Winter 2025/2026 health response plan. Callers describe their symptoms, software proposes a next step, and a clinician must approve the outcome before it takes effect. More than 36 thousand people had gone through this digital symptom-checking assistant by late December, the figure SPMS itself reports.
Steve Biko Academic Hospital — Complex paediatric neurosurgery, including the removal of brain tumours, now involves AI-assisted robotic surgery at Pretoria's Steve Biko Academic Hospital. Before operating, surgeons convert MRI and CT scans into detailed 3D reconstructions of the brain, then guide robotic arms through delicate movements while tracking progress continuously. Gauteng's health department has named the facility South Africa's flagship centre for public-sector medical innovation, driving broader adoption of robotic and AI-assisted surgery nationally.
Department of Health – Abu Dhabi — A virtual replica of Abu Dhabi's entire healthcare system now lets planners catch disease trends before they escalate. The Department of Health built the model, called Population Health Intelligence (PHI), together with Microsoft, folding in medical records alongside behavioural and ecological factors to support an AI population health platform that flags risks and simulates policy responses ahead of time. Obesity and cancer rank as its opening priorities, and officials call it the first effort of its kind at this scale, applying AI-driven public health management across the emirate.
Hospital E (Bệnh viện E) — Hospital E in Hanoi now runs an AI spectral CT diagnosis system, the Philips Spectral CT 7500, marking a first for northern Vietnam. Nearly 500 scans followed in month one, 45% coronary angiography, with its AI-integrated CT scanner telling benign cysts from malignant lesions at 96% accuracy.
An October 2024 Associated Press investigation, with research by Allison Koenecke (Cornell) and Mona Sloane (University of Virginia), found OpenAI's Whisper routinely hallucinated, inventing text no one spoke. Whisper underlies a Nabla ambient-documentation tool that 30,000-plus clinicians in 40 health systems used for some 7 million visits. The researchers judged nearly 40% of hallucinations harmful or concerning, among them a fabricated medication and invented commentary.
Tampa General Hospital — AI call handling at Tampa General Hospital has fielded more than 2 million calls since September 2025, trimming wait times 58% and lifting appointment capacity 21%. No vendor is named for the call-centre system, so this AI contact centre deployment rests on Tampa General's own account rather than an independent evaluation.
Ministry of Health, National Centre for Virtual Health — Oman's national AI diabetic retinopathy screening has examined over 43,000 retinal images and flagged over 12,000 patients, about 30% of those screened, for specialist review, per the director of the Ministry of Health's National Centre for Virtual Health. Over its first seven months it reached 91.3% accuracy and 85% sensitivity, cut specialist eye-clinic waiting lists 87%, reduced appointment waits from 62 days to eight and trimmed operating costs 43.2%. AI eye assessments now take roughly fifteen minutes, where pupil-dilating drops once meant about an hour and a half; phase one spans 25 health institutions.
Damascus University — Ada mirrored the prescribing doctor outright on 82% of 100 complaints and partly on 17% more, beating Symptomate's 68% and 25%, in AI symptom checker testing across five Damascus community pharmacies. Treatment advice matched the physician 49% of the time for Ada, 37% for Symptomate. Damascus University staff logged differing caution, Symptomate sending 53% of visitors to a doctor against Ada's 43%; the diagnostic app comparison stays an unrefereed preprint.
Hospital Escuela de Tegucigalpa — Hospital Escuela in Tegucigalpa reported 98% sensitivity for an AI sepsis detection system tested in 2026 on 100 emergency patients. The tool paired natural language processing with a wearable biosensor bracelet, reaching an F1-score of 0.848 and kappa=0.847 against clinical diagnosis. Researchers said AI-based early sepsis screening should support triage decisions, not replace clinical judgement.
Holy Family University, Batroun — AI in hospitals stopped predicting resilience once rival factors entered the equation — p = 0.143, despite a raw r = 0.340. Holy Family University in Batroun canvassed 412 healthcare professionals across Beirut and South Lebanon on enduring armed conflict. Crisis protocols, NGOs cooperation and security challenges survived instead; the authors credit institutional preparedness and joint working, rating artificial intelligence integration promising yet limited alone.
Canada Health Infoway — Canada Health Infoway's national AI Scribe Program shows AI medical scribes easing paperwork for most Canadian family doctors after six months. Among 12,019 clinicians enrolled June 2025 to February 2026, 70 percent cited lighter administrative strain and 80 percent cited reduced cognitive load from automated note-taking. Weekly hours saved averaged 1.11, same-day charting rose from 55.9% to nearly 75%, and monthly use of AI clinical documentation tools grew from 244,000 to 1.22 million encounters, per an independent review by the Centre for Digital Health Evaluation at the Women's College Hospital Institute for Health System Solutions and Virtual Care.
Syrian Private University — Gemini 1.5 Pro answered 86.1% correctly and ChatGPT-4o 60.8% when AI audiogram interpretation met specialist verdicts on pure-tone charts from 309 people aged six to 90. Syrian Private University clinicians ran the trial at Damascus Hospital, or Al-Mujtahid Hospital, submitting each chart as raw figures; 274 proved deaf to some degree. Cohen's kappa reached 0.848 for Gemini against 0.578 for ChatGPT, and machine hearing-loss triage belongs under supervision, never alone.
Mangochi District Hospital / Kamuzu University of Health Sciences — IMPALA, an AI patient monitoring platform in over 20 Malawian hospitals, cut child deaths 40 to 51 percent at two sites in 2023-2024 versus the prior year. It flags deterioration three hours early via nurse tablet alerts. Over 100,000 patients have used wards with the tool, now also in Tanzania, Rwanda, Kenya, Zimbabwe and Gabon, extending AI-powered patient deterioration prediction regionally.
Damascus University — Damascus University engineers built AI hysteroscopy image classification separating healthy endometrium from endometrial polyps and endometritis in fertility patients. Four transfer-learned architectures, VGG-16, VGG-19, DenseNet-121 and EfficientNet-B0, were fitted to 1500 annotated stills drawn from 200 procedures, scoring 85 to 89%. DenseNet-121 led at 93% sensitivity, 98.8% AUC on polyps, 90% precision on inflammation; the deep learning decision aid awaits outside datasets and pathology-confirmed labels.
Office of the Auditor General of Ontario — Every one of the 20 vendors cleared for Ontario's AI medical scribes procurement list failed to pass testing cleanly, Auditor General Shelley Spence's review found, turning up at least one inaccuracy apiece. Fabricated content, wrong entries and incomplete or missing information all surfaced during evaluation, with nothing to show that additional scrutiny had followed to lower the danger of invented output before rollout. Spence warned that unreliable AI-generated clinical documentation risks producing harmful or insufficient treatment decisions for patients throughout Canada.
In the MASAI randomised controlled trial of 105,934 Swedish women using the Transpara reading-support tool, AI mammography found 29% more cancers than standard double reading, cut radiologists' reading workload by ~44% and did not increase false positives (interim, Lancet Oncology 2023). Final results (The Lancet, 29 January 2026) add safety endpoints: ~12% fewer interval cancers (1.55 vs 1.76 per 1,000), 27% fewer aggressive (non-luminal-A) cancers and screen-detection sensitivity of 81% vs 74%.
In a cluster-randomised trial in Nature Medicine (2026), 'AI Consult', an LLM clinical decision-support tool in 16 Kenyan primary-care clinics (>9,600 patients), markedly improved documentation and treatment planning and cut antibiotic-related costs, but did not significantly change patient outcomes: 14-day treatment failure was 2.2% vs 2.0%, with no difference in hospitalisation, death or satisfaction. It is RCT-grade evidence of an LLM copilot improving clinician process without (yet) shifting hard outcomes.
AI-driven diagnostic error is the #1 patient-safety concern for 2026 in ECRI's annual Top 10 list (published March 2026), which cites inconsistent real-world performance. No major patient-safety authority had ranked AI first before.
A passive AI risk-trajectory display plus staff education failed to improve patient outcomes in a cluster-randomised trial (Scientific Reports, February 2026; 10,422 inpatient visits, 85-bed cardiology/cardiac-surgery ward). Deployed AI monitoring can fail to help, on trial-grade evidence.
Penda Health and Easy Clinic — AI Consult, Easy Clinic's tool for AI diagnostic support in Kenya, is a clinical decision support tool built into its EMR platform. It cut diagnostic errors 16% and treatment errors 13% across 40,000 visits at 15 clinics per a Penda Health study, with history-taking omissions down 32% and safety alerts declining as clinicians adapted. Published in Nature Medicine.
SPEC-AI Nigeria Investigators — Doctors running SPEC-AI Nigeria, a six-hospital randomised controlled trial, doubled peripartum case detection using AI-guided cardiomyopathy screening against routine obstetric care. The AI-enhanced digital stethoscope method required just 47 screenings per extra diagnosis uncovered. Results appearing in Frontiers in Cardiovascular Medicine mark Nigeria's inaugural randomised health AI trial.
A four-centre observational study in Poland (ACCEPT/OperA project), published in The Lancet Gastroenterology & Hepatology in August 2025, gave the first real-world evidence of AI deskilling among clinicians: endoscopists' adenoma detection rate on standard, non-AI colonoscopies fell from 28% before routine AI exposure to 22% after. The design was not randomised, and the authors note its population differs from the main ACCEPT trial target, so generalisability is unconfirmed. It is a rare documented downside of clinical AI.
OntarioMD — OntarioMD polled more than 1,600 Ontario family doctors, nurse practitioners and practice managers: 67% had used AI in care, measuring AI use among clinicians. AI scribes reached 58%, decision support 18% and inbox assistants 7%; 72% want provincial associations to vouch for safety. Fieldwork ran in March 2026.
Ministry of Health of Ukraine — Ukraine's Ministry of Health polled staff under the WINWIN 2030 digital development strategy: 80.5% of healthcare workers turn to AI in healthcare work, 21.6% habitually and 58.9% sporadically. Among medical professionals, 65.7% back medical AI if doctors retain the last word.
Only 1.6% of 256 healthcare staff surveyed across Kabul City hospitals and clinics showed good AI knowledge in healthcare, against 44.7% poor and 53.9% insufficient, with attitudes running ahead of understanding at 47.3% positive, 37.5% somewhat agreeing and 15.2% negative. The picture for AI in Afghan healthcare is limited knowledge but favourable attitudes: 82.0% had never attended an AI conference and 57.8% learned what they knew from social media. Fieldwork for the study, published in Annals of Medicine and Surgery, ran January to February 2025.
Ministry of Health (Türkiye) — Türkiye's Ministry of Health has run over 2 million results through AI breast screening since 2024 via its AI-supported Mammography Screening Reporting System. Images from KETEM cancer early-diagnosis centres and Healthy Living Centres flow into one central system, where AI triage pushes suspicious findings up radiologists' lists and those patients get priority appointments.
M&I/Partners — M&I/Partners' eighth AI Monitor, with replies from 52 Dutch hospitals, finds hospital AI use concentrated in care administration and billing (80%), then diagnostics (71%) and communication (65%). Most use is small-scale (53%) and 8% large-scale; 48% have a worked-out AI vision and strategy, and 76% an AI policy.
Damascus University — AI adoption in pharmacy practice reaches 55% of a 400-strong Syrian sample, halved between 200 dispensing chemists and 200 undergraduates, ChatGPT the best-known name at 53%. Practising chemists outscored undergraduates on knowledge, 5.7% versus 2.5%, yet undergraduates proved readier to adopt later; those dispensing under two or beyond six years used AI least. Writing from Damascus University, where curbed connectivity limits reach, the authors urge rules and tuition before everyday pharmacist AI use. Fieldwork ran September to November 2024.
Sudan Medical Specialization Board — AI adoption among surgeons in Sudan stays notional: of 185 surgery residents and specialists canvassed, 68.7% recognised artificial intelligence, 9.4% applied such tools on patients, 10.3% attained advanced literacy, 22.3% were formally schooled. Blockers: frail infrastructure 87.6%, absent tuition 79.1%, cost 72.4%; amid fighting, 92.5% lacked kit and 78.3% steady power or internet. Wanted clinical AI use: teaching 68.2%, perioperative decision-making 65.4%. Fieldwork spanned October 2024–June 2025, plus 20 interviews, at Khartoum's Department of General Surgery, Sudan Medical Specialization Board.
Reported AI use among U.S. physicians reached 81% in early 2026, up from 38% in 2023. It more than doubled in three years.
The FDA had authorised 1,524 AI-enabled medical devices by 30 March 2026, about 76% in radiology, up from 1,451 at end-2025 and 1,247 in mid-2025. The list grows every quarter, with another update in July 2026.
Nova Scotia Department of Health and Wellness — Over 440 medical professionals in Nova Scotia now use AI scribes to record patient visits, according to Department of Health and Wellness figures. Their licences come from Canada Health Infoway's national programme, which funds a free first-year AI documentation licence for eligible primary care clinicians; the province holds 500 funded licences, and Mika is its only approved scribe. Infoway says 95 per cent of AI-scribe users nationwide report less administrative burden.
Czech Ministry of Health — A Czech Ministry of Health analysis of 105 healthcare facilities across all regions puts AI adoption in Czech healthcare at 67.6% using or testing it, led by radiology at about 82.9%, then cardiology. The Czech-built Carebot chest X-ray tool, which highlights small lesions for less experienced doctors, runs at hospitals including Nový Jičín, Šumperk, Tábor, St Anne's in Brno, FN Bulovka and Slaný.
Italian public health system (multiple ASLs) — Roughly 14% of Italy's population, over 8.5 million people, now has access to Lunit's AI cancer screening technology for mammography and chest X-ray, deployed by more than 10 Local Health Authorities (ASLs) spanning Lazio, Lombardia and Veneto, the company reports. By December 2025, Lunit anticipates onboarding almost 4 million additional residents, expanding AI-based cancer detection to around 20% of the country. Lunit has not had these figures independently audited.
Ministry of Health (Vietnam) — AI training healthcare workers is the goal of a new Vietnamese scheme, AI Opportunities for Healthcare Workers, targeting 15,000 professionals by 2026. Google-certified staff lead five modules at Hanoi Medical University on prompt engineering and responsible AI use, feeding into digital transformation in healthcare nationwide.
US digital-health startups raised US$14.2 billion across 482 deals in full-year 2025, up 35% on 2024's restated US$10.5 billion, and AI-enabled companies took 54% of dollars, from 37% in 2024. By H1-2026 funding had reached US$7.4 billion across 244 deals, about US$1 billion ahead of H1-2025, with US$100M+ mega-deals taking 45% of capital.
Health Service Executive (HSE) — The Health Service Executive (HSE) issued guidance under Ireland AI regulation splitting health-service AI projects into five stages, from opportunity spotting to live oversight. An Opportunity Registration Platform, a Fundamental Rights Impact Assessment and an AI Inventory accompany the rules, and this clinical AI governance underpins the 2026 AI for Care strategy.
Medicines and Healthcare products Regulatory Agency — MHRA's National Commission into the Regulation of AI in Healthcare wants staged sign-off for new AI models, ongoing tracking of AI medical devices, open safety data, and tougher enforcement powers. Testimony from over 12,000 people across a year shaped this AI in healthcare regulation blueprint, aimed at medical AI systems oversight from approval onward.
National Commission into the Regulation of AI in Healthcare — UK AI regulation could adopt staged authorisations, with new AI models run under tight supervision before sign-off, if the MHRA heeds the National Commission into the Regulation of AI in Healthcare. NHS consultant Professor Alastair Denniston chaired the year-long inquiry, drawing on over 12,000 people. Continuous real-world surveillance, tougher enforcement and searchable safety data on AI-enabled medical devices and adverse incidents round out this healthcare AI rulebook.
Medicines and Healthcare products Regulatory Agency and Manchester University NHS Foundation Trust — The Medicines and Healthcare products Regulatory Agency (MHRA) and Manchester University NHS Foundation Trust (MFT) have opened the Manchester Sandbox, a testbed for AI medical devices running inside genuine NHS care under regulatory watch so real-world data builds beyond what trials alone show. Devices topping the list use AI to flag people whose long-term conditions carry a raised risk of complications, joined by wearables and ambient voice tech unveiled at the launch on MFT's Oxford Road Campus. Regulators, innovators and patient groups will later meet at a roundtable to choose which AI-enabled medical devices testbed projects come next.
Ministry of Health — Three risk tiers now govern AI language tools in healthcare across Kuwait, under a guide the Ministry of Health approved via Sheikh Dr Salman Al-Sabah. AI clinical decision support tools cannot replace human judgement for diagnosis or treatment, patient data cannot enter unapproved platforms, and AI-generated clinical notes need human checks first. Facilities must flag AI-linked incidents to regulators.
Ministry of Health, Ghana — Ghana's Ministry of Health is drafting a National Artificial Intelligence Health Policy and Guidelines: legal, ethical, operational healthcare AI rules. The healthcare AI policy spans governance, regulatory oversight, data protection, workforce readiness and digital infrastructure, aligned to WHO, OECD and EU standards. The health AI framework debuted at Ho's International Conference on AI in Healthcare and Pharma.
Health Information and Quality Authority (HIQA) — Ireland's Health Information and Quality Authority (HIQA) has published the first national guidance on safe, responsible AI in health and social care, commissioned by the Department of Health. New AI must be judged safe and trustworthy before use, and HIQA calls accountability the foundation, requiring clear accountability lines and robust governance for every AI use.
Všeobecná zdravotní pojišťovna ČR — Všeobecná zdravotní pojišťovna ČR has adopted rules deciding which novel technologies, among them AI in diagnostics, win a place in the reimbursement system, with trials opening 1 July 2026. VZP names three opening fields: chronic disease management, diagnostics, and backing prevention. A yet-to-be-named board of specialists will judge every application on safety, gain and value for money — machine learning reimbursement never having had an orderly Czech pathway, leaving products to creep in from below.
Ministry of Health and Family Welfare — SAHI, India's Strategy for Artificial Intelligence in Healthcare, hands hospitals, developers, researchers and policymakers one shared roadmap for steering AI in healthcare responsibly. Debuted at the India AI Impact Summit 2026, it demands tools clear safety, efficacy and ethical benchmarks before reaching patients, guiding medical AI adoption rather than dictating particular products.
Ministry of Health and Child Care — The Ministry of Health and Child Care will neither bring AI into public hospitals nor fully digitise their records, Deputy Minister Sleiman Timios Kwidini told parliament, fearing hackers could expose patient data. Answering Mbizo MP Corban Madzivanyika, Kwidini called AI "a very essential tool" in development, yet said Zimbabwe AI policy excludes hospitals. This health AI policy covers state-run facilities.
Western Cape Government — South Africa's Western Cape Government has adopted its first provincial AI healthcare policy, setting ethical rules for deploying AI in public health where resources run thin. It rests on five principles: human oversight, fairness, transparency, beneficence, non-maleficence. A Khayelitsha pilot informed it, testing AI-enabled retinal screening for diabetic retinopathy across four clinics.
South African Health Products Regulatory Authority (SAHPRA) — Selling an AI/ML-enabled medical device or in-vitro diagnostic in South Africa now needs SAHPRA clearance, per AI medical devices regulation dated 19 August 2025. Four risk tiers, Classes A to D, set evidence required for safety, effectiveness and post-launch monitoring under this AI/ML-enabled diagnostics oversight scheme. A Medical Device Establishment Licence is likewise compulsory for manufacturers and importers.
Health Professions Council of South Africa (HPCSA) — South Africa's Health Professions Council now requires doctors and nurses to tell patients whenever AI tools shape their treatment, under a refreshed Booklet 20: Ethical Guidelines on the Use of Artificial Intelligence. Practitioners carry full accountability for clinical outcomes no matter how much AI in clinical practice contributes, since the rules cast such systems as aids to human judgement, never substitutes for AI clinical decision support. The framework folds this obligation into the council's broader ethical code rather than carving out a distinct set of standards for artificial intelligence.
Department of Health, Medical Device Division — Hong Kong regulators set new evaluation rules for artificial intelligence tools used in diagnostic and treatment equipment under Technical Reference TR-008, effective from 3 January 2024. The Department of Health's Medical Device Division published the standard, which spells out listing criteria for AI medical devices within the Medical Device Administrative Control System. Machine learning-based products now face defined technical benchmarks for approval, marking a shift toward structured oversight of AI in healthcare technology.
Garante per la protezione dei dati personali — Ten binding rules now govern AI in healthcare within Italy's National Health Service, laid out in a Decalogue from the Italian Data Protection Authority (Garante Privacy), forming the country's public health service oversight standard. Designing, training and running these systems must satisfy three anchoring principles: openness about how automated choices get made, human checks on algorithmic outputs, and protection against biased or unfair treatment.
Ministry of Health of the United Republic of Tanzania — Systematic AI in healthcare integration across Tanzanian hospitals is the goal of the Policy Framework for Artificial Intelligence in Tanzania Health Sector, issued by the Ministry of Health of the United Republic of Tanzania. It sets out the technologies, safeguards and roles needed to bring adoption into everyday practice, lifting patient outcomes nationwide. Success depends on joint action by regulators, clinicians, academics and training bodies, building a base for future rules on AI health applications.
Tika Legal — Tika, from ex-reporter Alison Mau and lawyer Zoë Lawton, has opened publicly, pairing sexual-harm reporting with Tika Legal representation. Solicitors apply AI in legal services, as machine learning plus a large language model surface accounts possibly naming one alleged perpetrator. Staff vet matches before survivors decide, keeping AI-assisted legal support under human control.
Bradley — Bradley, a Birmingham, Alabama-based Am Law 100 firm, has made Legora its primary legal AI platform. The firmwide AI system puts models inside everyday tools for more than 1,200 attorneys and business professionals, COO Rance Sapen says.
Driven Lawyers — Amsterdam now hosts the Netherlands' first AI-native law firm, Driven Lawyers, started by Arvid van Oorschot and Jelle Drok. Software named Driven Drafter powers this AI in legal services practice, baking founders' frameworks into drafts and tying results to checkable sources. Clients pay flat fees reflecting automation savings, across disputes, IP and contracts.
Cooley — Cooley has released GO Public, a proprietary offering built by its capital markets lawyers and legal engineers and run inside ChatGPT Enterprise, bringing AI drafting for IPO filings to companies entering the public markets. A system of purpose-built AI agents handles the Form S-1, with Cooley as design partner to OpenAI. Cooley promises a first, bespoke S-1 in minutes not days, with the legal AI agents revising as transaction terms shift, under prompts and targeted review.
K&L Gates — K&L Gates has rolled out AltaClaro's DepoSim, an AI deposition simulator made with Verbit, across its Litigation practice, including Commercial, Labor & Employment and Intellectual Property litigation. One of six DepoSim Design Partner Program firms, K&L Gates beta-tested the deposition practice tool.
GESSEL — Harvey AI in legal document review is now firm-wide at GESSEL, a Polish law firm, woven into how it runs day to day. Beyond reviewing files, the system handles research, contract comparison and drafting, freeing senior staff for tasks needing real judgment. Marcin Macieszczak frames this AI-assisted contract drafting push as reshaping GESSEL's operations, not a bolt-on tech trial.
Morgan & Morgan — Morgan & Morgan, America's largest plaintiff firm, has launched MX2, a legal AI platform built in-house that has already handled medical-record requests in the hundreds of thousands and auto-retrieved many police reports. The firm will spend at least $1 billion on legal tech and AI over 10 years, adding to $300m of technology spending in the past five.
Grupo Financiero Inbursa — Grupo Financiero Inbursa, among Mexico's largest financial institutions, has rolled Harvey's legal AI out across its legal team. Inbursa's litigators use it for thousands of yearly regulatory and court notices, litigation analysis and oral-hearing preparation, and Harvey says this AI for lawyers will cover contract review, regulatory analysis and M&A due diligence as use spreads.
HSP Gruppe — German advisory firm HSP Gruppe has brought generative AI for advisers to around 900 staff via ChatGPT Enterprise, for research, advisory work, knowledge management and client communications. It claims over 40,000 hours freed a year, some 28,000 billable and worth about €3.8 million in potential revenue, with 98.6% of surveyed users more productive; no independent evaluation exists.
PwC Africa — PwC Africa has built an AI sustainability reporting tool that chief executive Dion Shango says shrinks two weeks of work to a couple of hours. He presented the reporting software at an African business AI summit as proof of AI moving from ambition into everyday operations.
Greenberg Traurig — Greenberg Traurig's roughly 3,200 lawyers now have the next-generation CoCounsel Legal, Thomson Reuters' agentic legal AI, for research, drafting and planning across each matter, plus a Brief Builder that converts research into draft court-ready briefs. The rollout follows four years of work with Thomson Reuters; outputs cite Westlaw and Practical Law content for lawyers to review.
BakerHostetler — BakerHostetler has launched Matter Research+, an AI-assisted research and knowledge tool co-developed with Clio, drawing on Clio's Vincent AI assistant. The tool sits in a new proprietary Practice Intelligence Center joining the firm's institutional knowledge to AI workflows.
Ghana Ministry of Finance — Ghana's Ministry of Finance has given its Legal Directorate Kwame AI's Eskwai system for AI legal research across Ghanaian case law and legislation, Kwame AI says. The directorate researches and advises on law for the ministry and other economy-managing state bodies, now backed by the legal AI tool.
Cleary Gottlieb, Freshfields, Weil, Williams & Connolly — Cleary Gottlieb, Freshfields, Weil, and Williams & Connolly are among law firms using Google Cloud's Gemini Enterprise for Legal. The legal AI platform handles tasks such as citation verification, Google Cloud says.
Nelson Mullins — Nelson Mullins is deploying Harvey's generative AI platform across every practice group. The firm counts more than 1,300 attorneys, policy advisors and other legal professionals.
Clifford Chance — Clifford Chance has launched an AI knowledge-management platform, built with Epiq Advisory for Law Firms and Microsoft on Microsoft 365 and Azure, to curate firm knowledge for tools like Microsoft Copilot. It holds knowledge as reusable 'skills' AI can read, since Copilot struggled to search the firm's iManage store.
Kinstellar — Kinstellar chose Laurel's AI timekeeping and work-intelligence platform to automate billing and time recording for about 400 lawyers in 15 offices across Central Asia, South-East Europe and Central and Eastern Europe. Timesheets should take 10 or 20 minutes instead of a few hours a day from early autumn 2026.
Davis Wright Tremaine — Davis Wright Tremaine has given the whole firm access to legal AI platform Harvey and Microsoft Copilot, plus Microsoft's AI Frontier Suite. The firm targets 90% adoption and top-quartile AI usage among law firms.
ArentFox Schiff — ArentFox Schiff has introduced 'FoxAI', which pairs the firm's proprietary tools with licensed third-party AI platforms. Firmwide training and a dedicated Innovation team come with it.
Goodwin Procter — Goodwin Procter has built an AI tool for venture-capital financing, made by six of its associates on a custom version of Anthropic's Claude. Firm chair Anthony McCusker told Bloomberg Law it plans to spend $25 million a year on such products.
Faegre Drinker — Faegre Drinker has moved past limited pilots to roll out Microsoft Copilot and the legal AI platform Harvey firm-wide. Chief technology and innovation officer Scott Angelo said curiosity predicted adoption better than practice area or age, and uptake sped up once senior partners started using the tools.
HEUKING — Licences for Harvey, the legal AI platform used by German commercial law firm HEUKING, have topped 1,000 firmwide. Weekly engagement is strong — 85% of monthly users return each week for AI in legal research, contract drafting and translation — a sixfold rise since the 50-person trial that started in July 2024. Attorneys still sign off on every output before use.
Haute Equipe — Haute Equipe has rolled out soev.ai, a Dutch sovereign-AI platform, across the firm. Every employee got AI-literacy training.
Ballard Spahr — Ballard Spahr has deployed Syllo's unified Litigation AI and Litigation Agent platform firm-wide, covering the litigation lifecycle from eDiscovery to trial for its 400-plus litigators. Rolled out after over two years of tests, it supports fixed fees per completed litigation task instead of billable hours, the firm says.
Liblaw — Liblaw has rolled out CJ Lewis for AI legal education in Liberia, the country's first legal-learning AI platform. Drawing on Liberian Supreme Court rulings from 1861 to 2026 alongside statutes and regulations, it turns complex law into plain explanations reachable by phone or computer for market women, farmers, students and journalists. Liblaw is Liberia's biggest digital law library.
Kuwait Bar Association — Kuwait Bar Association partnered with WKEEL.AI to launch the Wakeel AI legal research platform after more than three years in development. It equips the legal profession, government bodies and academic researchers with AI-powered legal drafting and analysis tools grounded in Kuwaiti statutes and judicial precedents.
CRIDO — CRIDO, Poland's largest law and advisory firm, now runs an AI workspace for lawyers powered by Wolters Kluwer's Libra platform. Staff handle case research, contract review and routine consulting through this AI-powered legal research tool, launched once an internal champions group had tested and compared competing options over several months.
Conselho Federal da OAB — Brazil's Federal Council of the Bar Association (OAB) launched OpenDetector, a free AI citation checker for registered lawyers, flagging invented legal provisions, wrong precedents and citations absent from official databases in AI-drafted petitions and opinions. The Forlex-built checking tool opens the OAB's national plan for AI in legal practice.
Johnson Stokes & Master — Johnson Stokes & Master runs internal AI agents on Microsoft Copilot under a governance-first structure, applying AI in legal services at the Hong Kong firm. Its Employment Legal Advice Copilot Agent handles high-volume employment advisory work, sorting matter context, flagging key issues, drafting first-pass output. Lawyers approve everything before clients see it, keeping Copilot-based legal document drafting under human sign-off.
WKB Lawyers — WKB Lawyers, an independent law firm in Poland, is giving all 160-plus of its lawyers access to Harvey's AI platform, a firmwide push into AI in legal services covering 40-plus practice areas from M&A to IP and financial rules. The software takes on due diligence, litigation support and compliance work once done by hand, freeing staff for higher-value advisory tasks, part of wider legal AI adoption.
FPS Rechtsanwaltsgesellschaft — FPS, a German commercial law firm, now runs Harvey, an AI platform built for legal work, speeding review of complex cases and bulky document sets. Staff pair this legal AI document review with AI-driven legal document analysis, and FPS says human lawyers still deliver every final legal call, not the software.
Bayerisches Staatsministerium der Justiz — Bavaria's Ministry of Justice built ALeKS (Anonymisierungs- und Leitsatzerstellungs-Kit) for AI anonymisation of court rulings, feeding 'Project 50k', a push to release 50,000 redacted Bavarian judgments over coming years. Munich Higher Regional Court piloted the system from 9 March 2026 before wider adoption followed. Central units still verify every machine-processed ruling by hand, keeping human review of automated judgment redaction in place.
Yes Architects — Yes Architects has AI architectural design live on two Azerbaijani building schemes, co-founder Yeşeren Ersan of the Turkish practice told the Invision Design Talks 2 forum. Roughly 20% of one scheme's drawings and 50% of the other's came out of machine-assisted design. Ersan puts today's saving near 10%, and sees 80-90% within five years.
Westfield Avocats — Rabat business law firm Westfield released Omniscient, an AI contract review platform trained on Moroccan law that checks agreements in under a minute and writes professionally worded amendments. Omniscient's contract-analysis AI pairs retrieval-augmented generation with lawyer-validated knowledge covering the Labour Code, Code of Obligations and Contracts, and company law.
Igazságügyi Minisztérium — Igazságügyi Minisztérium has switched on an AI legal research tool at jog.gov.hu, built across the whole body of Hungarian statute and every municipal decree. Justice minister Tuzson Bence framed the pitch plainly: ask in ordinary words rather than legal terminology, and the AI-powered legal search replies briefly, clearly and soundly in law. Practitioners and the public alike are the intended users, within the ministry's deregulation drive.
Legal Aid Commission, Ghana — Kwame AI has made Ghana's Legal Aid Commission its debut partner for Eskwai Pro Bono, opening a legal AI platform to 190 staff for research, drafting and file review. The aim: stretch 45 salaried lawyers across a country of over 30 million. It signals a fresh push to fuse free AI legal tools with public defender work nationwide.
Defensoria Pública do Estado do Rio de Janeiro (DPRJ) — Rio de Janeiro's Public Defender's Office launched VIA, an AI legal drafting platform producing petitions, analyses and documents from defenders' prompts inside the case system, from first client contact to court filing. The office says service could run up to five times faster, with humans reviewing all output.
Bowmans — Pan-African law firm Bowmans is rolling out Legora's legal AI platform after a successful pilot, Legora says. Bowmans lawyers now review documents faster, draft, and build and embed AI-driven workflows.
Corte Suprema de Justicia de la Nación — Argentina's Supreme Court has added AI case-law summaries to its advanced search of federal and national first- and second-instance judgments through an in-house viewer. Covering every LEX100 ruling, the plain-language texts may contain errors or biases and are no substitute for the ruling, the court warns.
The US Court of Appeals for the Sixth Circuit fined two attorneys $15,000 each in March 2026, plus the other side's appeal fees and double costs, for briefs with over two dozen fabricated citations. Calling it among the weightiest appellate sanctions yet over fake citations, the court based the penalty on lawyers' obligation to check every citation from any source, not on proof AI produced them.
Cabinet de Maître Raphaël Lapin — Raphaël Lapin, a Guadeloupe bar lawyer, applies AI in legal practice via dedicated legal software for case-law research, document review and drafting. A paid legal database, not general-purpose chatbots, drives this AI legal software. Verification and legal reasoning stay the lawyer's job.
Sam Okudzeto & Associates — One of Ghana's oldest legal practices, Sam Okudzeto & Associates, now handles research, drafting and document review through Kwame AI's Eskwai AI across its entire staff. Spanning over five decades, the historic outfit's move exemplifies AI adoption by law firm trends nationwide, pointing to broader momentum behind AI-assisted legal work.
Barreau de Paris (Paris Bar Association) — Barreau de Paris's entire workforce of 250 gains AI in legal services through Mistral AI's Le Chat, rolling out to a 100-strong test cohort from 18 December 2025 ahead of the wider January 2026 launch. Doctrine and Lefebvre-Dalloz previously equipped the bar with legal AI tools.
McMillan LLP — McMillan LLP now leads Canada as the first national firm to roll out Legora firmwide, opening the AI platform for lawyers to its entire staff. Selection followed scrutiny of security, privacy, dependability and features. That scrutiny underpins the firm's push into generative AI for lawyers.
HAVEL & PARTNERS — Legora, an AI platform for legal document work, has been brought into HAVEL & PARTNERS as the Czech firm looks to speed up drafting, review and analysis across bulky case files. HAVEL & PARTNERS, honoured as Innovative Firm of the Year 2025, previously relied on HP.AI and the WAIR research tool, and Legora now sits alongside them instead of taking their place. The addition widens the firm's toolkit for legal document automation.
Dalal Al Mulla Legal Group — Kuwait now has an AI legal platform, 'Musaed', unveiled by Dalal Al Mulla Legal Group at the NEXUS Exhibition. Al Mulla said the tool applies artificial intelligence legal consultations to speed up service delivery, sharpening accuracy while cutting turnaround time. The launch fits a wider push to modernise digital infrastructure and track shifts in legal-tech services.
RDJ LLP — RDJ is rolling out Legora, a generative AI platform for legal professionals, across its Cork, Dublin and Galway offices following a multi-practice-area pilot. Every trial participant said they would recommend the platform to a colleague. Lawyers at the Irish firm gain access to AI assistants for lawyers, tabular review and workflow automation.
Borden Ladner Gervais LLP (BLG) — Thomson Reuters' CoCounsel now runs firmwide at Borden Ladner Gervais LLP, giving staff AI legal research access. The tool covers generative AI legal drafting and document review, linked to Westlaw and Practical Law. BLG is one of two national full-service Canadian firms with this partnership.
Bahamas Industrial Tribunal — Bahamas Industrial Tribunal is trialling AI in judicial administration, using new software to speed legal research, draft judgment formats, and flag looming case deadlines, President Indira Demeritte-Francis confirmed. Chief Justice Ian Winder pushed for tougher safeguards to accompany this AI-assisted legal research drive.
Godoy — Godoy struck Latin America's first firmwide Harvey deal for AI in legal practice, rolling it out to every team. Staff now use it for research, reviewing documents and clearing repetitive work, marking a regional milestone for AI-assisted legal work.
Ministry of Law, Justice and Parliamentary Affairs, Government of Bangladesh — Daffodil International University and Daffodil Software Ltd built Bangladesh's e-Family Court platform free of charge to government. Dhaka Family Court-5 switched on first; Chattogram Family Court-1 followed on 30 November 2025, enabling online filing, digital records and partial remote hearings. About 250 to 300 judiciary, law ministry and Supreme Court staff piloted the family court digitisation system beforehand.
BSP (Bonn Steichen & Partners) — Passing through a rigorous selection process and successful pilot, Legora, a generative AI platform for legal professionals, is now live across every part of BSP (Bonn Steichen & Partners) in Luxembourg. Shifting lawyers away from routine tasks toward higher-value strategic advisory work drove the firm's embrace of AI in legal services. Managing partner Alain Steichen framed the change as vital for guarding BSP's competitive standing and sustaining client-service quality, calling generative AI adoption in law firms a long-term strategic priority rather than a quick efficiency fix.
Fidacta — Fidacta's AI deployment for notaries in the Netherlands, co-created with six Dutch notary firms, cuts administrative workload while keeping client data on domestic servers. Built on sovereign infrastructure from Nebul rather than ChatGPT or other hyperscale providers, the platform deploys AI agents that draft concept deeds, generate email replies and search knowledge collections, citing sources for each answer. The aim is easing paperwork burdens for the notarial profession while preserving confidentiality, part of a wider shift toward AI-assisted legal document preparation.
Cederquist — Weekly AI adoption in law firms hit 90% at Sweden's Cederquist through a strategic alliance with Vesence. Attorneys co-built the software across 2025, targeting AI-assisted legal deliverables that get reviewed and quality-checked before use. It now runs daily inside M&A deals, letting staff structure documents, verify quality and strip out friction on projects.
Holding Redlich — Australian commercial law firm Holding Redlich now runs Lexis Create+ nationwide for AI legal drafting, an AI-powered legal drafting tool inside Microsoft Word. Staff locate clauses and precedents by merging LexisNexis databases with in-house knowledge sources in one document window. Phased training accompanied the rollout.
Anderson Mori & Tomotsune (AMT) — Anderson Mori & Tomotsune, a leading Japanese law firm, rolled out Bengo4.com's Legal Brain Agent, an AI agent for legal research, firmwide from November 2025 after a trial confirmed accuracy. This legal-specialist AI agent draws its output exclusively from statutes, case law, guidelines and expert reference works, attaching citations that can be verified instantly, sidestepping fabricated references typical of general-purpose LLMs.
Consejo General del Poder Judicial — KENDOJ, the judicial software suite run by Spain's General Council of the Judiciary (CGPJ), now offers AI for judges via two additions: an automatic summary generator and a conversational assistant. The Centre for Judicial Documentation (CENDOJ) built both tools to run in a secure environment, turning court rulings into structured digests and letting magistrates query case files in plain language instead of manual searches. This cuts review time, part of a wider shift toward AI-assisted judicial document analysis in Spain.
MAQS Advokatbyrå — All 180 employees at MAQS Advokatbyrå now use AI in legal services following a firm-wide rollout of Legora's platform, expanding beyond an earlier trial. Staff across the Swedish firm's Stockholm, Gothenburg, Malmö and Sundsvall locations gained shared access to the legal research tool for reviewing and organising large volumes of case material. MAQS is ranked among Sweden's leading full-service commercial law practices.
Al Tamimi & Company — Legora's AI platform is now live across Al Tamimi & Company, the biggest full-service firm spanning the Middle East and North Africa, in the region's opening rollout of Arabic legal AI. Every lawyer gains research, drafting and review capabilities through a custom Arabic interface the pair engineered jointly. Headquartered in Dubai, the firm becomes a frontrunner embedding Arabic-language AI within legal practice.
Sheppard Mullin — Sheppard Mullin has handed Legora to over 1,200 attorneys firmwide for AI legal research, diligence, drafting and negotiations. A head-to-head pilot preceded the move, feeding the Am Law 100 firm's wider innovation strategy. Playbooks and certification aim to embed AI in legal practice.
JDP Drapała Chomiuk & Partners — Legora AI now handles document review at JDP Drapała Chomiuk & Partners, a Polish business law firm: hundreds of files scanned under oversight, extra arguments surfaced, expertise shared firm-wide. This is AI in legal services support, not substitution. A security audit and AI-fundamentals training preceded the shift toward legal document analysis using generative AI.
Seitz — Seitz, a German commercial law firm, rolled out AI document analysis in law firms to roughly 150 lawyers after comparing rival AI tools. The Legora platform enables AI-assisted document review, data extraction and legal research firmwide. Seitz said the benefit to everyday casework was obvious from the outset.
Ministry of Justice of Kazakhstan — Kazakhstan plans an AI smart assistant lawmaking tool to help draft regulatory legal acts, said Deputy Minister Bekbolat Moldabekov of the Ministry of Justice on 23 October 2025. It will apply international best practice, run legal monitoring, and flag clauses clashing with Constitutional Court rulings — an AI-assisted legal drafting step forward. This follows Digital Bailiff, live since August 2024, credited with saving citizens over 1.9 billion tenge (US$3.5 million) in enforcement fees, per the ministry.
Dentons — Drafting, research and document review at Dentons' New Zealand offices now run through Legora, a specialist generative AI platform marking the firm's entry into AI in legal practice. Charles Spillane described it as a foundational move toward broader legal document review software adoption.
White & Case — White & Case now runs Atlas firmwide, its first conversational AI assistant law firm application built on the firm's own proprietary AI platform. Deloitte helped build the system on Microsoft Azure, giving lawyers web lookup, saved chats, ready-made prompts, file uploads and the ability to write custom prompts. Rather than depend on one supplier, the firm treats Atlas as a base layer, adding outside AI tools through a buy-and-build approach for its in-house legal AI system.
Herzog Fox & Neeman — Herzog Fox & Neeman now runs Harvey AI firm-wide across its Technology Group, targeting cross-border tech M&A work. Lawyers use the platform for accuracy and speed across multi-jurisdictional deals, an AI legal tech deployment building on Herzog's earlier push into AI-assisted legal services.
Ro Sommernes — Ro Sommernes now runs Legora across the whole organisation, following a trial that convinced leadership to expand this legal AI tool company-wide. Erik Sandtrø, Managing Partner, judged it the safest and most fitting choice; Lynne Skajaa praised the AI platform for legal work as a joint human-machine effort lifting counsel quality, easing cooperation and adding client value at the Norwegian firm.
Cliffe Dekker Hofmeyr (CDH) — Cliffe Dekker Hofmeyr's AI legal software deployment spans generative AI legal platforms Harvey and Legora, picked so lawyers get the better-suited system per task. The South African firm trialled both with over 186 staff, then added compulsory ethics coaching and human oversight once licences went through.
Allen & Gledhill — Allen & Gledhill, among Singapore's biggest firms, built A&GEL, a bespoke platform for generative AI in law firms that drafts case summaries, automates document reviews and flags clauses non-compliant with statutes and earlier cases. The legal GenAI pilot, open to paralegals through directors across banking, corporate real estate, maritime and other practice groups, runs on-premise protecting client confidentiality.
MinterEllison — MinterEllison has adopted Legora firmwide, weaving AI in legal services into core workflows such as due diligence after a thorough selection process. This large Australian law firm now handles matters through one connected system instead of scattered tools, aiming to extend artificial intelligence for law firm processes across most operations.
Ministry of Digital Development and Transport and Ministry of Justice — E-qanun.ai now lets Azerbaijan's legal community mine the country's legislative database through one AI legal search platform. Built by the National Artificial Intelligence Center, the system reads queries by keyword, by meaning, or through a blended method, then hands back AI-written summaries instead of raw statute text. The launch pushes legal research automation forward, aiming for faster answers for practitioners and broader public reach into the law.
Ministerio Público de la Provincia de Buenos Aires — Buenos Aires province's Procuración General has confined all generative AI public prosecutor's office activity to one channel: ChatIA, cleared under Resolution PG 1475/25. Five prosecutor and defender offices now rely on it to write, revise and shorten legal filings, rather than turning to any outside AI legal document drafting tool. Entry demands a signed ethics pledge, sits behind encryption, and every user's actions are logged.
Wikborg Rein — Legora's AI legal platform now runs firmwide at Wikborg Rein after a successful pilot. Staff gain shared tools for live cross-document teamwork, ready-made expert prompt sets, and analysis of bulky spreadsheet data. Geir Sviggum said it lifts research input quality and speeds standardisation of everyday legal software tasks.
Goodwin Procter — Legora, an AI platform for legal document work, now runs across Goodwin's entire 1,600-lawyer firm, built with input from Legora's own AI team. Lawyers use it to summarise files, flag issues, pull contract terms, draft documents and review litigation material, applied uniformly across offices and departments worldwide. Support came through targeted training sessions and recurring open office hours, embedding generative AI in law firms among staff practising different specialities.
Domański Zakrzewski Palinka (DZP) — Domański Zakrzewski Palinka (DZP), a Polish law firm, ranks among the earliest strategic adopters of Legora, an AI platform for legal teams, advancing AI in legal services locally. Staff use it to sift paperwork, gauge exposure and reach findings quicker, boosting legal AI adoption. DZP says the software aids, not replaces, lawyers.
Allens — Allens becomes the first Australian law firm to partner with Legora, giving lawyers firm-wide access to this AI in legal services tool for tailored client collaboration. Thomson Reuters CoCounsel now handles contract review, agreement summarisation and legal research across the firm, marking an AI-driven legal research and drafting rollout.
Singapore Academy of Law — About 10,000 LawNet subscribers, over three-quarters of Singapore's private lawyers, now have access to GPT-Legal Q&A, a new AI legal research engine unveiled by the Singapore Academy of Law. Rather than searching by keyword, users pose questions in plain language and get answers grounded directly in local statutes and court judgments. The underlying model draws on a dataset built entirely from Singapore's own legal materials — judgments, law reports and legislation — with contract law as its first focus before wider legal question-answering coverage follows. TechLaw.Fest hosted the unveiling.
Clayton Utz — Clayton Utz has gone firmwide with Harvey, past a trial of 100-plus lawyers—an instance of AI for law firms. It covers drafting, due diligence, extraction, summarising and contract checks. It joins Harvey's earliest Australian clients, holding data in-country under a new secure setup, lifting legal AI tools uptake locally.
PwC Belgium — PwC Belgium has given every tax professional access to an AI tax assistant, the Tax AI Assistant, built with Harvey and trained specifically on Belgian case law and legislation. Rather than producing generic output, the tool drafts tax content tailored to domestic regulatory requirements. The rollout forms part of a broader ambition to position the firm as a leader in generative AI for tax advisory within professional services.
Szecskay Law Firm — Hungarian firm Szecskay Ügyvédi Iroda has started piloting AI legal software built by Icelandic developer Jónsbók, folding the tool into everyday casework and returning feedback so the system can be tuned for local practice. Dated 28 August 2025, the deal turns the independent Hungarian firm into a testing partner for AI in legal practice, joining an existing client base of government bodies, other legal practices, banks and publicly traded firms operating in various countries.
Wardyński i Wspólnicy — Legora, an AI platform for lawyers, now runs at Wardyński i Wspólnicy, the first Polish firm to embrace the tool. A 2024 trial with over 70 staff across all departments led to full use for drafting, review and analysis. Legal AI is spreading across Poland's legal sector as machine learning takes on routine work.
Webber Wentzel — Webber Wentzel now runs Legora across its practice, deploying AI in legal services after a twelve-month pilot by the South African firm. Legora becomes the debut anchor technology inside Webber Wentzel's Innovation Lab, where the pair will jointly craft tools for Africa's legal sector.
Olaniwun Ajayi — Nigeria's Olaniwun Ajayi now pioneers Harvey AI legal work, the country's and West Africa's first adopter of a generative platform built solely on legal data, advancing AI in law locally. Wired into Thomson Reuters, Westlaw and LexisNexis, it speeds research, drafting and case review while keeping client secrecy and data privacy intact.
Tilleke & Gibbins — Tilleke & Gibbins, spanning Thailand and five neighbouring countries, has adopted Harvey, an enterprise-grade AI platform for the legal industry. This AI in legal services tool speeds drafting, research and file review. Custom agentic workflows for Southeast Asian legal markets follow next, widening legal AI adoption.
Tilleke & Gibbins — Harvey, an enterprise-grade legal AI platform, now runs firm-wide at Tilleke & Gibbins, extending AI in legal services across Southeast Asia. Teams in Cambodia, Indonesia, Laos, Myanmar, Thailand and Vietnam use it for drafting, research, document review. Custom agentic workflows for legal practice are planned, tailored to core practices and Southeast Asian markets.
Latham & Watkins — Latham & Watkins has extended Harvey's generative AI suite firmwide, giving every lawyer worldwide access under one licence — a major push into AI in legal practice. Uses span research, drafting and document analysis. Sustained generative AI investment stays central to the firm's strategy, making this legal AI adoption permanent, not experimental.
KMSC Law — Over 100 KMSC Law staff spanning legal, administrative and operational roles now apply AI legal research via Alexi's legal intelligence platform, deployed firm-wide within a private cloud environment. Uses include drafting, summarisation and correspondence, with nearly 70 per cent of these users non-lawyers.
Mishcon de Reya — Mishcon de Reya has switched on Legora, an AI platform for lawyers, firmwide after a three-month trial. Staff at all levels used it to extract data, summarise files, draft correspondence and check work. This follows the firm's earlier build of deReyAI, a generative AI tool for legal professionals.
Polsinelli — Am Law 100 firm Polsinelli now gives roughly 1,200 attorneys nationwide an AI legal assistant, Thomson Reuters CoCounsel Core, covering transactional, regulatory and litigation work. Adoption reaches 70 percent of associates and 50 percent of shareholders. CoCounsel Drafting nears completion, extending Polsinelli's AI in law firms strategy.
Bell Gully — Bell Gully now ranks as New Zealand's first major law firm to embed AI legal research firm-wide, having struck a licence for the Harvey platform. Over 200 lawyers gain access for case lookup, storing files and tougher matters, per Andrew Petersen, framing it as a step in generative AI adoption for legal work.
Banning Advocaten — A successful pilot led Banning Advocaten to extend the Zeno AI platform firmwide. The Dutch firm now uses it for AI-assisted legal research, document review and drafting. Leadership calls this push into AI-assisted legal services a careful, responsible step to lift quality and speed.
Supreme Court of Korea — A W14.5 billion contract has gone to a KT-led consortium tasked with building South Korea's Supreme Court AI platform for judges, an AI-assisted judicial platform designed to accelerate legal research, pull case issues automatically and support judgment drafting. KT's proprietary legal-specialised language model, layered atop Mideum 2.0, drives the roughly four-year build, joined by partners Conan Technology, Elbox and MT Data. Shrinking case backlogs and sharper judicial output count among the hoped-for gains, per the court's own internal forecast rather than any independently verified result.
Gowling WLG — Gowling WLG has become the first Canadian law firm to deploy Harvey AI legal software nationwide, extending it to every practice group. Personnel now rely on the generative platform for research, diligence checks, reporting and client advisory needs. Beforehand, an internal audit had already surfaced hundreds of promising uses for AI in legal services.
Poder Judicial de Chile (Chilean Judicial Branch) — Chile's Judicial Branch switched on an AI-powered legal document search tool, the Jurisprudential Search Engine, unlocking meaning-based lookup across more than 1.5 million rulings instead of keyword matching. An AI legal research tool now filters by court, commune, legal norms, case category and timeframe, pairing alerts with saved queries via Clave Única digital identity login. Ministers Adelita Ravanales and María Cristina Gajardo led the two-and-a-half-year build.
Consejo de la Magistratura de la Nación (Poder Judicial de la Nación) — Argentina's Consejo de la Magistratura has given formal backing to a judicial AI assistant, powered by OpenAI's GPT-4.1-nano, through Resolution 1486/2025 covering the federal judiciary, Poder Judicial de la Nación. Inside the case portal, Portal de Gestión de Causas, lawyers, experts and prosecutors will get help resolving everyday queries once Phase 1 begins testing on 29 August 2025. Staff serving the courts internally join later, in Phase 2, only after officials assess how that trial run performs — an early instance of generative AI reaching case-management systems at court level.
University of Cienfuegos / National Organisation of Collective Law Offices — Cienfuegos lawyers now trial Alic.IA, an AI legal assistant from University of Cienfuegos academics, ahead of expansion across Cuba. It works as both a chatbot for legal practice queries on Cuban criminal law and a system drafting over 60 document types, approved after debuting at the Abogacía 2025 congress in March, following a pilot since late 2024.
HPP Attorneys — HPP Attorneys, a Finnish law firm, now runs the Legora AI platform firmwide after a successful six-week trial. This AI in legal services deployment aims to shift 16+ hours monthly per lawyer from routine to client-facing work, part of a wider push for AI adoption in law firms.
Ultima Ratio — Ultima Ratio, a Polish arbitration platform, cut justification drafting from three hours to seconds — a 98% jump — using AI arbitration awards software running OpenAI on Microsoft Azure. It checks facts against common-court and Supreme Court precedent to draft a ruling, which arbitrators revise as an AI-generated draft ruling before finalising.
Guilex Avocats — KYTISA, a new AI legal research in Guinea platform, was unveiled in Conakry by Guilex Avocats, a Guinean firm counted among Africa's top 20 legal practices. Subscribers gain a single hub for domestic statutes, judicial rulings and legal news, alongside an AI-powered chatbot built to automate routine legal tasks. Offered in French, English and Spanish, the launch marks the firm's push into an AI-driven legal research tool aimed at reaching users well beyond one language group.
Quinz — Quinz, a Belgian firm, applies AI in legal practice, rolling out Legora's platform for firm-wide document review, search and summarisation. Teams scanned over 300 Belgian Supreme Court rulings against tight deadlines and shrank a competition-probe review from weeks to days via AI-assisted legal research.
Arendt & Medernach — Arendt & Medernach, Luxembourg's largest independent law firm, has embedded generative AI in legal practice through Legora's platform, now standard across daily casework. A December 2024 pilot logged over 8,000 queries and 30,000+ data points, clearing the way for full rollout. Staff apply it to data extraction, prospectus review, compliance checks and AI-powered legal research.
Shardul Amarchand Mangaldas & Co — Shardul Amarchand Mangaldas & Co (SAM) becomes India's first law firm to deploy generative AI in legal services firm-wide, rolling out Harvey's complete toolset across seven offices after a year-long trial. Lawyers now use these AI-powered legal work tools for contracts, due diligence, research, forecasting and client pitches.
Plesner — Plesner, a Danish law firm, now ranks among the first in its market to install an AI governance platform law firm use requires, picking GRACE from 2021.AI to oversee generative AI across the business. Every employee learned to apply the tools safely and effectively. Several language models run inside one governed structure, with uptake, per Plesner, fast and broad, supporting responsible generative AI use firm-wide.
King & Wood Mallesons Australia — King & Wood Mallesons has scaled AI in law firms to more than 1,200 users spanning Australia and Singapore, building on a six-week pilot involving about 200 lawyers. Harvey, a generative AI tool for legal document drafting, now also checks compliance, condenses bulky files and proofreads.
Dentons — Every employee at Dentons across Europe and Central Asia now has access to DAISY, a homegrown AI in legal services platform offering secure generative capabilities. Prague-based staff under Břetislav Šimral, part of the firm's Europe Innovation and Intelligence unit, designed the system. Its debut version pairs a protected chat function with preset options for condensing and composing documents, complemented by a language-translation feature and a searchable internal policy guide.
Poder Judicial de Chile — Chile's Judicial Branch is bringing AI transcription in labour courts to the Labour Court of Arica, the first labour tribunal nationally to gain the capability, via the Justicia Ágil, Transparente y Automatizada (Justa) project. Minutes from hearings get produced without manual input, trimming paperwork and speeding rulings. Mulchén and Antofagasta ran earlier trials of automated minute-taking, but Arica marks Justa's debut in a labour setting.
ENS (Edward Nathan Sonnenbergs) — ENS counts among the first African law firms to embed Harvey AI at full scale, applying the legal AI platform to contract review, due diligence, compliance and litigation. Automation handles routine document tasks while lawyers retain final sign-off, echoing a continental shift toward AI in legal services.
Supreme Court of Ukraine — Ukraine's Supreme Court now uses AI legal research to scan full-text rulings and generate short legal summaries in seconds, via the new Database of Legal Positions 2.0. ChatGPT and Anthropic models power the tool after training on hundreds of prior cases, linking to the Unified State Register of Court Decisions and the E-Court System. No registration is needed, making this generative AI for case-law analysis a rare open example inside a national judiciary.
Rechtbank Rotterdam — Rechtbank Rotterdam turned to generative AI for help writing a criminal verdict, an apparent generative AI criminal verdict first, restricting the tool solely to wording the sentencing-motivation passage — this AI-assisted judicial drafting stayed out of evidence weighing or actual rulings, with staff verifying each suggestion. Privacy risks led the court to reject mainstream AI options and instead push for a dedicated, secure internal system.
Ministry of Justice — Four modules make up Digitalny Asystent Sędziego, an AI judicial assistant unveiled 19 March 2025 by Poland's Ministry of Justice under the Cyfrowy Sąd (Digital Court) programme: filing digitisation, a CJEU case-law knowledge base, automatic claim calculation for foreign-currency mortgage disputes, and algorithm-generated draft rulings. Poznań is among the first three district courts to get the tool, with the wider AI-powered judge's assistant scheme due by mid-2026.
Guangdong Baocheng (Futian) Law Firm — Guangdong Baocheng (Futian) Law Firm has deployed Fa Xiao Qi for AI contract review, case analysis and document drafting, built on DeepSeek's large model as a legal AI assistant. Drawing on over 1 million case records, it boosts clause-comparison accuracy by 80% and drafts compliant contracts instantly, underpinned by a specialised legal model and knowledge base.
Beijing Dacheng (Kunming) Law Firm — AI in legal practice has moved on-premises at Beijing Dacheng (Kunming) Law Firm, which installed DeepSeek-R1-671B, DeepSeek-V3-671B, QwQ-32b and GLM4-9b on its own servers rather than the cloud in March 2025. Contract review, legal research and drafting tasks get routed automatically to whichever model fits best, with KIMI and Tongyi Qianwen brought in alongside the locally hosted systems for added coverage. Housing everything on Uptime Tier III certified infrastructure keeps client information private and under domestic control, a priority driving broader adoption of AI in law firms wary of external hosting.
Castrén & Snellman — Castrén & Snellman, based in Helsinki, became Finland's first law firm to adopt Harvey AI in legal services, following a successful trial. The domain-specific tool handles routine tasks, freeing lawyers for advisory work while preserving personalised client service — an early case of law firm AI adoption in the Nordic region.
Ballard Spahr — Ask Ellis anchors Ballard Spahr's AI legal work suite: a chatbot honouring a co-founder, built on OpenAI models in a private tenant for data security. Companion legal AI tools include Ballard X-Ray, a searchable document repository with chat, and Know Your Client, blending internal records with outside sources for client metrics. A technology innovation director led the build.
Filip & Company — Filip & Company has become Romania's first law firm to deploy Harvey, a generative AI system built specifically for legal work, and the first among independent firms anywhere in Central and Eastern Europe. This marks an early instance of AI adoption in law firms spreading beyond global players into smaller regional practices. Rather than drafting from scratch, lawyers now lean on this generative AI system for legal work for research, translation, contract review and regulatory checks, shifting their time toward strategic advice instead of routine tasks.
Howard Kennedy — Jylo, a GenAI marketplace and document review tool, is live at Howard Kennedy, a UK top-100 firm, after a lengthy trial. Teams use it for AI document review, building playbooks that flag risk and spot clauses, with strong uptake. This contract review automation lets groups shape their own workflows.
Bruun & Hjejle — Danish law firm Bruun & Hjejle now embeds Harvey AI legal work in daily operations, speeding routine tasks, deepening research and untangling dense filings. Partners call this one strand of a deliberate AI strategy for stronger client service, part of a wider push toward AI-assisted document analysis.
Lietuvos Aukščiausiasis Teismas (Supreme Court of Lithuania) — Lithuania's Supreme Court is testing TeDIA, a pilot assistant using AI press release drafting for courts to turn rulings into draft statements. Staff then rework the machine-generated text with queries and manual edits before sign-off. The trial, running through mid-2025, extends 2024 groundwork with Vilnius University into practical AI-assisted court communications.
Interius Advocaten — Interius Advocaten scrapped its former systems for Innoverius, a Belgian legal AI suite built by Gert Loterman on Odoo, Microsoft 365 and OpenAI. Document drafting, case management and billing now run through this single legal practice management tool. Interius is the first firm using it.
Gide Loyrette Nouel — Gide Loyrette Nouel's AI tool for legal services, the Jimini AI legal assistant, now runs firm-wide after a pilot involving nearly 100 lawyers and staff across its offices in France and abroad. Over 60 lawyers took part before the 14 October 2025 confirmation.
N. Dowuona & Company — Ghana's first firm-wide AI legal assistant deployment: N. Dowuona & Company has given lawyers Eskwai Firm Plans, an AI-powered legal research tool from Kwame AI, for drafting and contract review. Kwame AI calls the firm a top practice, citing a month's trial before launch. No earlier firm bought Eskwai's plan.
Iranian Judiciary — Voice or text queries now reach RoboLegal, Iran's first AI legal robot, mining a database of Iranian statutes and rulings. Built for people priced out of representation or far from urban courts, it debuted in Tehran. Judges and lawyers also gain a tool for routine paperwork, cutting delays through AI in legal services.
Sejong Law Firm — South Korean law firm Sejong, ranked among the country's top five, now applies AI legal research via Harvey AI, an OpenAI-based legal generative AI tool, in a pilot running since April 2025 for chosen advisory matters. Overseas assignments involving bulky document sets rely on it most: hunting for precedents and regulations, drafting contract wording, and condensing due diligence files. This counts among the earliest legal AI research assistant rollouts in South Korea.
State Great Khural of Mongolia — TUS Solution LLC built the AI legal analysis module now running on Mongolia's D-Parliament platform, operated by the State Great Khural, to scan bills clause by clause for conflicts with constitutional human rights guarantees. Flagged provisions go back for revision, giving citizens clearer access to legal information. This is AI-assisted legislative review reaching a national parliament.
PHH Rechtsanwält:innen — Firmadigital.at lets founders complete AI GmbH incorporation entirely online, a first for Austria, built by Vienna-based PHH Rechtsanwält:innen with Infinity Vertigo. Users submit data and AI drafts the articles of association, while the platform also handles identity checks, shareholder agreement, notarial deed and commercial register entry remotely. Piloting preceded the rollout of this AI-supported company formation platform.
Barreau de Bordeaux — 2,200 France-based lawyers at the Bordeaux Bar, the country's fifth-largest, now rely on Haiku, a homegrown AI legal assistant, for drafting, research and document review, live since 1 September 2025 under a July deal. This reflects growing automation in legal services.
Ashurst ran firm-wide generative-AI trials from November 2023 to March 2024 with 411 partners, lawyers and staff in 23 offices across 14 countries. It reported first-draft time savings of about 80% on UK corporate filings, 59% on sector research reports and 45% on legal briefings, while stressing expert human review, as its lawyers still beat the AI on accuracy. Ashurst's own figures lack independent evaluation.
HSBC's Dynamic Risk Assessment, an AI anti-money-laundering system co-developed with Google, screens around 980 million monthly transactions. Group head of financial crime Jennifer Calvery says it catches two to four times as much financial crime as its rules-based predecessor, with 60% fewer false-positive alerts, freeing investigators for real risk. HSBC's figures lack independent evaluation.
District of Columbia Court of Appeals — The District of Columbia Court of Appeals on 3 September struck a Deutsche Bank unit's whole appellate brief in a foreclosure case for citing four non-existent authorities, and referred the matter to the DC Bar's Office of Disciplinary Counsel. The filer said Google's generative AI tool produced the citations; the three-judge panel held every firm lawyer who signed the brief shares some responsibility, while stressing it was not discouraging AI use.
Estonian translation market — Rates for Estonia's two thousand translation livelihoods slid from 18 euros in 2019 to under eight in tenders, the sharpest sign of AI in the translation profession. Demand shrank as machines took the craft, though the European Commission rendered 100,000-plus pages into the language last year with 55 staff. Triin van Doorslaer of Tallinn University, Nele Nikopensius at Luisa Keelelahendused and Elis Paemurd at the Commission's office report the squeeze on machine translation work.
Over 2,000 court decisions worldwide now record a party relying on AI-fabricated citations or authorities by September 2026, up from roughly 1,600 mid-year, at an estimated five to eight a day. Its maintainer admits the count understates the true number.
In 2026, 50% of lawyers called AI a major threat to the ban on unauthorised practice of law, up from 36% in 2025. It is a profession-specific fear of AI encroaching on regulated legal work.
High Court of Australia — Judges act as "human filters" for machine-generated legal arguments, Stephen Gageler, Chief Justice of Australia's High Court, told the Australian Legal Convention, warning that AI in courts poses "existential" risks. Machine content has exceeded sustainable limits, he said, since AI's rapid advance outpaces the judiciary's grasp of its dangers. His remarks flag mounting pressure from AI-assisted litigation.
Tribunal de Primera Instancia de San Juan — María V. Irizarry Centeno and Anissa Michelle Bonilla Irizarry face a $1,000 penalty from Judge Larissa N. Ortiz Modestti, who caught six bogus Supreme Court precedents fabricated within their filing through AI-generated fake legal citations. Officials at the Tribunal de Primera Instancia de San Juan in Puerto Rico describe it as the territory's inaugural sanction targeting AI hallucinated case law inside legal proceedings. The Supreme Court will decide if additional punishment against the attorneys is merited.
Ministerio de Justicia / JUFEJUS / UBA-IALAB — 5,000 members of Argentina's judiciary gave 81% approval to the Programa Piloto de Uso Estratégico y Responsable de IA Generativa en la Justicia, the Ministry of Justice's UBA-IALAB-run trial of AI in Argentina's judiciary. Drafting led the 3,000 prompts across 1,007 test bases (three each), at 30.7%, ahead of analysing information (17.3%) and case-law search (11.7%). Courts in Mendoza, San Juan, Tierra del Fuego, Misiones, Tucumán, Buenos Aires and CABA took part in this use of generative AI in the justice sector.
OAB São Paulo — The 2026 Report on the Impact of AI on Law, presented at the São Paulo bar association (OAB SP), puts frequent AI use among Brazilian lawyers at 77%, up from 55% a year earlier. Legal AI adoption was measured across more than 1,800 professionals nationwide.
Active generative-AI use in professional-services organisations (legal, tax, and risk and compliance) reached 40% in 2026, from 22% in 2025. Thomson Reuters' annual survey covers ~2,000 professionals.
Agentic AI is only starting in legal work. In 2026 about 15% of organisations have adopted an agentic AI tool and under 20% report widespread use; roughly half are planning or considering one.
Lawsnote — Nearly all legal professionals in Taiwan now use AI in legal work: Lawsnote's survey of 721 respondents put usage at 85.6%, about triple North America's 25-31% rate, reflecting strong AI adoption among legal professionals there. Employers, though, remain far behind: only 34.1% have formally sanctioned any tool, so 51.5% of staff turn to unapproved "shadow AI" instead. Among those relying on it, 43.4% cited efficiency gains above 40%.
Russian legal departments (Techologies of Trust / PravoTech survey) — AI in legal departments across the Russian Federation has reached 49%, up sharply from 19% just a year earlier in 2024, Technologies of Trust found in its benchmark of the legal function. Globally the picture runs ahead: the Association of Corporate Counsel puts worldwide AI adoption in the legal function at 52% for 2025. Spending lags the enthusiasm — merely 24% of departments have paid for AI tools at all, and three-quarters of those, 76%, kept outlay below 5 million rubles.
Russian judicial system — More than a third of Russian courts, 35%, now rely on AI in Russian courts, Viktor Momotov of the Council of Judges reports. Bashkortostan, Tatarstan, Vladimir and Irkutsk lead with speech-to-text for hearing transcripts. Komi's Supreme Court mines case law digitally, Vologda drafts rulings and reviews documents by machine, and Amur automates offence-protocol paperwork — signalling wider judicial AI adoption nationwide.
By mid-2026, 43% of law firms and legal departments had adopted generative AI, from 14% at the start of 2024, as experimentation became integration. This is organisation-level adoption; the earlier 14%->26% figure tracked individual professionals.
By September 2025, 61% of UK lawyers used generative AI in day-to-day work. In January, AI use stood at 46%.
Among US lawyers, reported AI-based tool use hit 30% in 2024, up from 11% in 2023. Large firms led: 46% at firms of 100+ attorneys.
18% of AML and compliance professionals have AI or machine learning in production. Another 18% are piloting and 25% plan to adopt within 12-18 months.
Lebanese Association of Certified Public Accountants — The Lebanese Association of Certified Public Accountants ran an AI qualification for auditors in Beirut, 8-11 September 2026, with the Arab Federation of Accountants and Auditors. Kamal Shehade, Minister of State for Administrative Development and Artificial Intelligence, gave his patronage. The Jordanian Association of Certified Public Accountants attended the artificial intelligence diploma opening.
Minnesota Supreme Court — Minnesota attorney Faisal S. Ahmed accepted a 30-day suspension of his law licence after a 2025 brief he filed cited non-existent court cases produced by an AI tool. Several Minnesota matters involve AI-fabricated citations, including briefs with unverified invented authorities from attorneys David Lutz and Frederic Knaak. The suspension goes further than the fines and sanctions courts had imposed for such conduct.
Ministry of Justice — Japan's Ministry of Justice on 21 August 2026 set conditions for permitted AI in legal services under Article 72 of the Attorneys Act, easing the unauthorised-practice bar. Contract review support now qualifies, alongside two other named uses, provided the tool skips litigation design and builds in misuse safeguards for customers — a shift toward sanctioned use of artificial intelligence in legal work. This clears a compliance path for vendors, leaving courtroom work to licensed lawyers.
Supreme Court of Guam — Guam's Supreme Court fined attorney Rachel Taimanao-Ayuyu $500 for AI use in court filing that left a brief riddled with errors. Chief Justice Katherine Maraman and Associate Justice Robert Torres found the document unmoored from existing law, faulting Taimanao-Ayuyu for skipping a thorough check of the machine-drafted text. She admitted leaning on generative AI without confirming its accuracy, and now follows a written policy on generative AI use in legal filings alongside continuing legal education on ethical AI practice.
Supreme Court of Guam — A $500 sanction hit attorney Michael Phillips over AI hallucinations in legal brief work before the Supreme Court of Guam. Chief Justice Katherine Maraman found he skipped the required check under the Judiciary of Guam's AI policy, which permits AI use but mandates verifying unverified AI-generated content in court filings first. The case involved the Guam Legislature's dispute over the governor's power to establish hospitals.
Supreme Court of Nepal — Nepal's Supreme Court has adopted the Judiciary Information and Communication Technology Policy 2026, permitting AI in courts for legal research, document handling and administrative support. Judgment drafting and writ petitions requiring judicial discretion stay off-limits. Systems must first clear checks for reliability, transparency and bias, reflecting a cautious stance on courtroom AI adoption.
Bar Association of Sri Lanka (BASL) — Sri Lanka's Bar Association has drafted rules for generative AI in legal practice, guiding lawyers who use such systems for research, drafting, case analysis and client contact. Confidentiality comes first: public AI platforms require lawyers to shield client data, stay answerable for AI-assisted output, and verify results before relying on them. Non-binding for now, the framework is billed as a living document set to evolve with the technology, an early marker of AI in the legal profession.
Corte Suprema de Chile — Chile's Supreme Court suspended a litigating lawyer for one month and fined her 5 UTM over AI-invented citations: her cassation appeal quoted doctrine from non-existent books generated by an AI chatbot. Its Third Chamber found a breach of procedural good faith.
Consejo de Estado — Colombia's Council of State warned sharply over AI-fabricated citations after finding that rulings a lawyer cited in a rejected jurisprudence-unification appeal were non-existent or mismatched. It ruled AI admissible only when decisions come from the judge's own reasoning, never deciding, weighing evidence or interpreting law, and referred the lawyer to disciplinary authorities.
Judiciary of Kenya — Fabricated citations from AI in court proceedings have pushed Kenya's Judiciary to draft a Judiciary Artificial Intelligence Policy setting practice directions for lawyers and judges. Under the proposed rules, anyone submitting AI-assisted filings must verify content beforehand and disclose the tools used, with applications graded by risk level. Public comment is open before the policy takes final form, addressing cases where AI-generated legal documents contained invented case references.
Board of Grievances of Saudi Arabia — Saudi Arabia's Board of Grievances has issued mandatory ethical guidelines and regulatory controls for AI use in judicial administration, covering development, deployment and operation of such systems across the body and its affiliated courts. All staff using these tools must comply, formalising AI governance in courts previously left to general guidance.
Corte Suprema de Justicia (Sala Civil) — Colombia's Supreme Court has fined a lawyer 15 monthly minimum wages over AI-invented case citations, its first sanction for AI misuse, in Auto AC739-2026 of its Civil Chamber. His appeal cited rulings absent from official repositories; challenged, he supplied four further fictitious ones, and admitted using an AI tool.
Qatar International Court and Dispute Resolution Centre — The Qatar International Court and Dispute Resolution Centre's procedural directive sets AI rules for lawyers and litigants appearing before its courts and tribunals from the start of 2026. Representatives must check AI-assisted material, flag AI-generated content in documents and declare AI-generated video or audio evidence or risk its exclusion. The AI directive followed a lawyer's contempt finding for citing AI-invented cases.
Qatar International Court and Dispute Resolution Centre — A Qatari tribunal disciplined a Dubai attorney whose filings cited AI-generated fake cases, invented rulings in Jonathan David Sheppard v Jillion LLC. Judges branded this intentional contempt, not sloppy research — a regional first for MENA. Punishment: a reprimand, with future use of fabricated legal citations to be named publicly.
Civil Court of Thailand — Thailand's Civil Court now requires lawyers and litigants to flag any AI-generated material used in pleadings and other court filings, under a regulation covering AI in legal filings. Disclosed sections must carry prescribed wording plus a certification confirming their accuracy, and failure to comply can be treated as contempt of court. The measure follows guidance the Supreme Court's President issued in September 2025, reflecting an effort to protect reliability and public trust in civil proceedings, and mirrors broader court rules on generative AI disclosure in legal documents taking shape elsewhere.
Law Society of Ireland — Narrow, low-risk jobs only: the Law Society of Ireland's stance on generative AI in legal practice, greenlighting admin help, summarising and checklists through ChatGPT-style software. Legal advice, document review, and citing case law stay off-limits for LLMs, with AI tools for law firms unsuitable beyond those limited tasks.
Judicial Council of Ireland — Chief Justice Donal O'Donnell's committee has issued Ireland's first AI guidelines for judges: administrative tasks and speeches are fine, but generative AI use in the judiciary excludes legal research or case analysis. Fabricated citations and bias are key dangers. Chatbot answers need independent verification before an Irish judge relies on them.
Supreme Court of The Bahamas — Lawyers appearing before The Bahamas' Supreme Court must now flag any generative AI in court proceedings, under Practice Direction No. 3 of 2025, issued 5 November 2025. The rule obliges them to check every fact and citation an AI tool produces and to stand fully behind whatever they file, regardless of how it was drafted. Three principles anchor the guidance: openness about AI use in court submissions, careful checking of outputs, and personal accountability for filings.
Workplace Relations Commission — Ireland's Workplace Relations Commission has issued AI guidance for litigants warning that generative AI tools risk fabricating claims about Irish employment and equality law. Filers stay fully answerable for whatever a submission says, no matter which tool drafted it, under this generative AI in legal submissions policy.
Corte di Cassazione — A judge's evaluative judgement cannot be delegated to AI, the Supreme Court of Cassation held (ruling n. 34481), since dependence on such systems erodes impartiality in AI in legal proceedings. TAR Lombardia separately penalised a lawyer whose AI-drafted appeal cited invented cases, referring the matter for discipline. Both rulings bound AI use by courts and legal professionals well short of replacing human judgement.
Supreme Court of Thailand — Barring artificial intelligence from weighing disputed facts or shaping judicial independence, Thailand's Supreme Court President has laid down binding guidance for AI in court case management. Judges keep sole authority to decide cases, and any staff turning to AI use in courts for administrative work must personally check the results are correct. The guidance further bans entering official secrets or personal data into such systems.
Consiglio Superiore della Magistratura — Italy's judiciary now has formal limits on AI use in Italian courts, set on 8 October 2025 by the Consiglio Superiore della Magistratura (CSM), the self-governing body overseeing magistrates. Court staff may lean on artificial intelligence in court administration for organising files, researching case law and producing standard draft formats, but evidence assessment, rulings and final decisions must stay human-made. CSM frames these limits as temporary, lasting only until the EU AI Act becomes fully binding in August 2026.
Supreme Court of Judicature of Jamaica — A ban on machine-drafted affidavits and witness statements now sits alongside mandatory disclosure duties under Practice Direction No. 1 of 2025, Bryan Sykes's new framework for generative AI in courts across Jamaica's Supreme, Revenue, High and Gun Court divisions. Lawyers and litigants who breach these terms face cost orders or contempt findings. Sykes presents the crackdown as reinforcing integrity rather than resisting judicial AI adoption.
Ministry of Law — Public comment is open from Singapore's Ministry of Law on a draft Guide for Using Generative AI in the Legal Sector, covering ethics, confidentiality and openness for lawyers. The Singapore Academy of Law, Law Society of Singapore and firms contributed, easing safer uptake of GenAI industry-wide. It sits beside IMDA's Model AI Governance Framework and the Supreme Court's guidance for court users.
Senior Courts of Belize — Belize now regulates AI use in courtrooms through Practice Direction No. 18 of 2025, issued by the country's Senior Courts. The rules require checking machine-produced text for accuracy, flagging its role in filings, and keeping final decisions with people rather than software. Belize thus became one of the first Commonwealth Caribbean nations to set formal limits on generative AI within its justice system.
Bermuda Supreme Court — A landmark ruling has legitimised AI e-discovery: Bermuda's Supreme Court, in Fourworld Global Opportunities Fund Ltd v Enstar Group Limited, dismissed manual review as financially impractical for modern disputes. Justices found generative AI paired with forensic software sharply cuts disclosure costs and burden, among the earliest common law judgments to sanction such AI-assisted disclosure tools.
Supreme Court of the Turks and Caicos Islands — Practice Direction 1 of 2025 brings an AI regulation for courts to the Turks and Caicos Islands, binding lawyers, clerks and unrepresented parties under one Supreme Court rule. Drafting and research assisted by generative AI need checking and disclosure, but affidavits and witness statements stay off-limits. Privileged material must avoid public AI systems in favour of enterprise-grade platforms — court use of generative AI kept within firm limits.
Legal Practice Council (LPC) — Fake AI-generated precedent showing up in two South African court proceedings has pushed the Legal Practice Council toward drafting formal rules on AI use in legal practice. Academic institutions and IT specialists are helping the LPC shape standards that carry real teeth, since breaches could lead to disbarment. The council's response signals mounting alarm over generative AI tools inside courtrooms.
Tribunal Supremo de Puerto Rico — Puerto Rico's Tribunal Supremo has written artificial intelligence into new Professional Conduct Rules, adding Rule 1.19 on technological competence and diligence to govern AI use in legal practice. Lawyers must grasp what AI tools can and cannot safely do, check machine-generated material with care, and never file AI-produced court documents unless a human has independently confirmed them first. Effective 1 January 2026, the measure makes Puerto Rico the first US jurisdiction with a freestanding ethics rule of this kind.
Iraqi Judiciary (Karkh Misdemeanor Court) — Iraq's judiciary now treats AI misuse fraud as an existing crime rather than a legal gap, despite no Iraqi statute naming artificial intelligence outright. Baghdad's Karkh Misdemeanor Court says schemes built on AI, whether forgery, defamation or blackmail, still trigger standard penalties under current law. Meanwhile the Rusafa Court of Appeal's Deputy Chief Judge pointed to a live prosecution under Article 456, evidence that judges are already folding AI-related crimes into the Penal Code rather than waiting on new rules.
Procuración General de la Nación (Federal Prosecutor's Office) — Argentina's Procuración General de la Nación (Federal Prosecutor's Office) launched an AI strategy programme, the Programa de Inteligencia Artificial del Ministerio Público Fiscal, under Resolution PGN 14/25. Duties span ethical and technical guidance, project review and staff training, an AI governance framework referencing the EU AI Act, the Council of Europe's Framework Convention and UNESCO recommendations.
National Chamber of Legal Advisers (KIRP) — Poland's National Chamber of Legal Advisers (KIRP) unveiled AI tools legal advisers can now follow as official ethics recommendations, the country's first such framework, on 13 May 2025. Microsoft Poland and law firm Sołtysiński Kawecki & Szlęzak helped shape the rules, which tackle client confidentiality, data protection, who bears responsibility for content an AI system produces, and how to bring these systems into daily practice safely. KIRP called it AI ethics guidance for the legal profession, a living text meant to change alongside shifting rules and emerging technology.
Federal Court of Australia — Federal Court of Australia guidance now covers generative AI in legal practice: Chief Justice Debra Mortimer's Notice to the Profession follows a 28 March 2025 statement flagging the issue. A dedicated 'AI Project Group' will consult key legal personnel while drafting formal AI rules for filings. Lawyers stay fully liable for submitted material regardless of what tool assisted.
Supreme Court of Pakistan — Justice Syed Mansoor Ali Shah's 18-page opinion for Pakistan's Supreme Court, dated 11 April 2025, demands formal rules covering AI in courts before judges lean further on systems like ChatGPT or DeepSeek. Machines, the bench warned, can widen judicial capacity yet never substitute for human reasoning, discretion and empathy. Drafting duties now fall to the National Judicial Policy Making Committee alongside the Law and Justice Commission, tasked with defining what AI use in the judicial system should look like. Separately, the Federal Judicial Academy's 'Judge-GPT' is already deployed, backing roughly 1,500 district judges.
Conselho Superior da Magistratura (CSM) — Human rulings cannot be handed off to machines, Portugal's Conselho Superior da Magistratura (CSM) has decided, restricting AI in judicial activity to a supporting role only. Under the new guidelines, evaluating proof and interpreting statutes remain tasks for the judge alone, with the judicial governance body also setting up monitoring mechanisms covering how courts deploy AI tools for judges.
Consejo de la Magistratura de la Ciudad Autónoma de Buenos Aires — AI use in judiciary now falls under enforceable limits in Buenos Aires: Resolution 206/2025 (26 February) bans handing rulings to machines. Its Guide of Recommendations and Directives sorts systems by risk tier and demands data-protection reviews before high-risk deployment. Wherever AI-assisted court proceedings happen, officials must keep audit logs and disclose the technology's role.
Caribbean Court of Justice — Barbados's CCJ court published Practice Direction No. 1 of 2025, setting rules for generative AI in court proceedings across the Caribbean Court of Justice's jurisdiction. Attorneys, parties, witnesses and self-represented litigants must verify GenAI-generated citations and meet accuracy and professional responsibility standards under the new framework. The step gives judges an early template for handling AI-assisted filings region-wide.
Caribbean Court of Justice — Binding guidance on generative AI in court proceedings now covers lawyers, litigants, witnesses and unrepresented parties throughout the region, under Practice Direction No. 1 of 2025 from the Caribbean Court of Justice. Saint Lucia falls under this GenAI regime too, since it belongs to CARICOM and sits within the CCJ's appellate reach. Instead of a single-country rule, this judicial AI regulation stretches across the wider Caribbean bench.
Supreme Court of NSW — A new NSW Supreme Court block on generative AI in legal evidence covers character references, affidavits and witness statements, binding lawyers, litigants and judges alike. Machine-written reasons and judgments are barred too. Submissions may still incorporate AI-drafted court documents, provided citations are checked independently.
Orde van Vlaamse Balies and Ordre des Barreaux Francophones et Germanophone — Avocats.be and the OVB, Belgium's French-speaking and Flemish bar associations, issued joint rules on AI for legal work, treating such tools as optional rather than banned or compulsory. Lawyers must verify machine-generated outputs, protect client confidentiality, pseudonymise personal data, and stay fully accountable for finished work product. The framework leaves adoption of AI in legal practice to individual judgement while fixing firm duties on verification and data handling.
Grand Court of the Cayman Islands — Personal accountability now falls on lawyers who lean on artificial intelligence while drafting court filings, following Justice Asif KC's decision in Bradley & anor v Frye-Chaikin. This marks the first written guidance from the Grand Court of the Cayman Islands on AI in legal submissions: using such technology to speed up litigation draws no criticism, provided the person relying on it checks outputs for fabricated citations or other invented content before filing. Attorneys who skip that verification could face costs sanctions, and rival counsel now carry a duty to warn the bench whenever an opponent's AI tools in legal proceedings risk misleading it.
Bundesrechtsanwaltskammer (BRAK) — Confidentiality, transparency and professional duty frame fresh rules on AI use in law firms from Germany's Federal Bar Association (BRAK), set against the EU AI Act. Risks named include hallucination and skewed training data, yet the lawyer alone bears responsibility for AI-assisted legal work. BRAK calls it advisory, not a substitute for individual judgement.
Slovenská advokátska komora (SAK) — Slovakia's SAK enforces binding AI use in legal practice rules, launched via Resolution No. 12/4/2025 of 31 March, sharpened by a November 2025 revision. Lawyers must pass an ethical check before feeding confidential client data into AI tools, under standards covering secrecy, protection and conduct. New guidance sets baseline requirements and template disclosure wording for clients on AI-assisted legal work.
Court Artificial Intelligence Research Society — Judges behind South Korea's Court Artificial Intelligence Research Society drafted "Guidelines on the Use of Artificial Intelligence in the Judiciary", a five-chapter framework addressing AI use in the judiciary. Bias and hallucination risks for judges applying such tools sit alongside rules on adopting the technology and duties for litigants to flag artificial intelligence in judicial work whenever they file submissions.
Nigerian Bar Association Section on Legal Practice — Nigeria's legal sector gained formal guidance on AI in legal practice through the Guidelines for the Use of Artificial Intelligence in the Legal Profession in Nigeria, issued by the Nigerian Bar Association Section on Legal Practice. The framework equips lawyers with ethical expectations covering client confidentiality, sound professional judgment, safeguarding of data and management of risk when deploying such tools. It emerged alongside other technology-focused releases unveiled at the association's Annual General Meeting, marking a coordinated push toward artificial intelligence oversight within the country's bar.
Israel Bar Association National Ethics Committee — Israel's Bar Association has mapped out boundaries for AI use in legal practice, telling practitioners to double-check every AI-generated answer before relying on it. Client secrets stay out of open-access systems unless the client signs off in writing, and practices are steered toward closed, in-house platforms instead of public tools when building legal AI applications.
Woori Asset Management — Woori Asset Management rolled out WISH (Woori Investment & Strategy Hub), giving every employee AI agents in asset management over one consolidated data store. Equity, fixed-income, alternatives and compliance teams pose plain-language questions, getting AI analysis for fund teams that speeds reporting and control checks.
Biess (Banco del Instituto Ecuatoriano de Seguridad Social) — ABI, a generative AI virtual agent for loan enquiries, answers Biess members on mortgage, pawn and unsecured loans and other products in a renewed call centre. Biess is the social security bank; calls pass to human staff as needed.
Barclays — Barclays is extending Anthropic's Claude bank-wide, a step for AI coding in banking. Targets: 50% of developers on Claude Code by end-2026, most software engineers in 2027, widening AI-assisted software engineering. Over 16,000 colleagues already consult a Claude knowledge assistant; Global Markets sorts roughly 120,000 client emails daily.
DOM.RF Bank — DOM.RF Bank cut staff effort roughly a third using AI agents in banking to vet escrow-based individual housing construction contracts and track project-finance spending. Agents also support IT development, analytics and paperwork, and AI contract review ended waits for loan tranches.
Millennium Management — Millennium Management has deployed about 1,600 AI agents for hedge fund staff, dubbed "digital twins", and is widening access to more than 7,000 employees, investment professionals included. Each twin holds a separate identity, email address, permission set and audit trail, covering research, meeting preparation and inbox triage at the $97bn firm. Of more than 150 pilot users, 97% engaged daily; new joiners get these personal AI assistants automatically.
Yucho Bank — Yucho Bank (Japan Post Bank) gives its corporate sales division UPWARD, an AI agent producing AI sales meeting summaries through voice analysis of in-person visits. UPWARD spots visits via smartphone location and files uniform call reports in the CRM system. Yucho Bank credits the AI-generated notes with lighter admin, better proposals.
HSBC — HSBC unveiled HSBCnio at Sibos 2026, adding AI in corporate banking for institutional and corporate clients. Client-side AI assistants reach authorised transaction and account information via HSBC's Model Context Protocol.
Techcom Securities (TCBS) — TCBS, a Vietnamese securities firm, adopted agentic AI coding through AWS's Kiro, used by more than 460 staff. All engineers use Kiro, plus non-technical staff.
BankIslami Pakistan — BankIslami has switched on TezRaftar, a home-built platform bringing AI credit underwriting to consumer loan origination. Car, housing, personal and state-scheme lending now share one digital process and branch-wide rules, an automated loan assessment trimming manual checks and delays.
Société Générale — Société Générale has bound itself to Anthropic in an agreement it bills a decisive move towards corporate AI in banking, promising sharper output and better customer treatment. Group-wide moves should yield savings of 500 million to 600 million euros by 2029, 350 million already pinned down. The lender set it out in a roadmap from Paris on 21 September 2026, its firmest word on enterprise AI adoption.
Pensions Awareness Ireland — Pensions Awareness Ireland has released Ali, billed as the country's first AI retirement coach, to mark Pensions Awareness Week 2026. Ali takes a saver's age, chosen finishing age, gross yearly pay and monthly payments, then returns a dashboard setting out projected income. Schooled on the Irish pensions rules, the tool replies from that dashboard and shows whether a pot will fund the later life wanted — AI pension planning without an appointment.
PureHealth — PureHealth has put AI in health insurance to work at The National Insurance Company, Daman, aiming it at 60 million documents a year and cover for 3 million people in the UAE. Better than 80 per cent of that paperwork should land as structured digital records, trimming turnaround from days to minutes. The artificial intelligence document processing chain should also cut manual rework by over 50 per cent, freeing Daman staff for fraud checks, clinical oversight and member support.
Mosaic Insurance — Mosaic Insurance has switched on Halo, AI underwriting for SMEs quoting, binding and issuing eligible policies in minutes. Submissions arrive auto-structured and enriched; tricky risks go to a human underwriter. North American wholesale brokers get this automated small-business underwriting first, starting with SME cyber.
Mercado Pago — Mercado Pago, the digital account of Mercado Libre, has brought Mago to Chile — AI in personal banking that settles routine money chores by conversation: spending, transfers, bills, benefits. Speech, typing or a shared photo all serve as input, and nothing executes before the user confirms. Conversational finance totals over two million exchanges, near 224,000 monthly.
Mosaic Insurance — Mosaic Insurance has stood up an in-house AI Lab, pairing four engineers with business teams so AI in underwriting stops stalling at pilot stage and reaches daily work. Capability ships across claims, finance and operations, engineers sitting beside colleagues to spot problems and push fixes live. Machine-assisted risk assessment shapes how the insurer reserves and decides, by its own account.
Prudential Hong Kong — Prudential Hong Kong has given every financial consultant AI Underwriter, built with Alibaba Cloud, cutting preliminary AI insurance underwriting checks from several days to minutes. Prudential reports accuracy above 95% and hallucination rates under 2%. Human underwriters keep the final decision; the automated triage supports rather than replaces them.
HSBC Malaysia — HSBC Malaysia gave relationship managers Wealth Intelligence, generative AI for wealth management condensing Chief Investment Office views, market notes and unit trust fund analysis. Linda Yip, who leads international wealth and premier banking, said the adviser copilot adds speed; trust stays human.
Prudential Hong Kong — Prudential Hong Kong's AI Underwriter, built with Alibaba Cloud, is now fully live, bringing AI insurance underwriting to financial consultants at the point of sale. A chatbot returns initial guidance in minutes, not days, using customers' financial, medical, occupational and residential details. Human underwriters keep final say on automated underwriting advice.
Nationwide Building Society — Microsoft 365 Copilot now serves 1,000 finance staff at Nationwide Building Society, alongside 2,500 engineers on GitHub Copilot, under an organisation-wide rollout. 15 million customer records have also moved onto one cloud data platform, advancing AI in financial services and a broader banking AI transformation.
KIWI Finance — KIWI Finance has handed roughly 400 loan intermediaries agentic AI in lending, running FlowX.AI tools inside a sealed environment. One conversational agent halves onboarding for newcomers; an eligibility module doubles response speed for customers. Romania's biggest credit brokerage thus joins State Street, Banca Transilvania, OTP Group and UniCredit as a FlowX.AI production user of AI-assisted brokerage.
Korea Life Insurance Association — The Korea Life Insurance Association has introduced AI complaint handling alongside AI review of advertising and internal regulations. Speech-to-text transcribes consumer–counsellor calls live, classifies each complaint type and gives counsellors supporting material mid-call. Three applications went live this month, and the association aims for 17 projects, including translation, contract review and legal review, during 2027.
MUFG Bank — Japan's Mitsubishi UFJ Bank deployed Omakase AI, a ZEALS multimodal AI agent platform, for AI customer service in banking, first fielding account opening campaign queries by natural voice or text. Hallucination detection and range control keep this bank customer support AI within financial institution security and accuracy standards.
M&T Bank — M&T Bank has given AI copilots to over 15,000 of its roughly 22,000 employees, extending Microsoft Copilot, GitLab code generation and an internal repository called Edison to software development, report drafting, call-centre analysis, identifying customer needs and flagging risk. Chief data officer Andrew Foster said AI call summaries save about six minutes each, and that the bank's data-lineage work on Solidatus and Monte Carlo predates generative AI. Employees must review all AI-assisted work.
United Overseas Bank — United Overseas Bank has put more than 300 AI and analytics use cases into production under an AI assurance framework that tests, validates and continuously monitors models across their lifecycle. It also runs 30-plus domain-specific chatbots and has given Microsoft Copilot to over 30,000 employees. Alvin Eng, UOB's head of enterprise AI, says scaling AI safely hinges on operating discipline, governance and controls more than on having the strongest individual models.
Agrobank — Agrobank now uses AI and data analytics in lending, internal controls and credit risk assessment. Internal auditor Shahob Eddin Masharipov says AI credit scoring on big data, drawn from tax reporting, telecom data and transaction activity, rates retail and small-business borrowers. Automated data checks cut microloan decisions to a few minutes with no credit officer involved, the bank says.
Woori Bank — Woori Bank has hired KT to rebuild its chatbot and advisory bot as AI agents in banking, linked to its 'AI Banker' financial consultation service through KT's Agent Connector so customer context follows across channels. KT has started development, aiming at sharper intent recognition and AI in outbound calls and branch consultations.
Grupo Aval — Grupo Aval has moved over 36 AI agents in banking into production in 18 months via ADL Digital Lab and its RE initiative, handling digital-channel data management and conversion for over three million Colombians. Technology vice-president Ernesto Gutiérrez gave the figures; Lulo Bank described AI spanning decision-making, risk management, operational efficiency and customer interaction.
Discovery Bank — Discovery Bank now handles over half of client interactions through its in-app AI banking channel, which works in natural language, chief executive Hylton Kallner told an African business AI summit. He said the in-app assistant was delivering the bank's highest customer-satisfaction scores yet.
IndiaFirst Life — IndiaFirst Life has chosen Salesforce Agentforce to place AI agents in insurance work across claims processing, customer service and new-business underwriting. The agents review KYC and medical documents, sort and answer service cases and check claims paperwork.
Experian — Experian has launched a credit-cards app inside ChatGPT, taking its credit-card marketplace into conversational AI. Consumers can compare and find card offers without leaving their ChatGPT conversation.
Bank of Ghana — The Bank of Ghana is applying AI, big data and machine learning to quicker economic intelligence, AI inflation forecasting, growth projections and earlier financial-risk spotting. First Deputy Governor Dr Zakari Mumuni outlined it at the Ghana Statistical Association's 4th Annual Statistics and Data Science Conference in Tamale. Fuzzy definitions, mismatched categories or badly checked data, he cautioned, can mislead even the most advanced forecasting models.
BBVA — BBVA's Brújula, an AI sales assistant for bankers built by its South American teams, helps corporate and institutional relationship managers in Colombia, Argentina, Venezuela, Peru and Uruguay rank opportunities, prep meetings, log client visits and draft financing proposals. Each interaction feeds the next client conversation, and the manager keeps professional judgement and accountability for decisions and the portfolio.
Banco Azteca — Banco Azteca, among Mexico's biggest financial institutions, has shifted its automation onto UiPath's Maestro platform, using agentic AI in banking to coordinate people, AI agents and systems across entire business processes. More than 300 production automations span customer service, HR, procurement, legal, fraud prevention and compliance, plus over 8,800 automated processes UiPath equates to more than 3,300 full-time employees' work, unverified by any independent evaluation.
Deutsche Bank — Germany's Deutsche Bank, a design partner for the Financial Research agent in Google Cloud's Gemini Enterprise for Financial Services, is rolling the AI research agent out across its Corporate Bank. The tool runs research end to end with source citations and audit trails, drawing on 50-plus specialised financial skills.
CME Group — CME Group is among the first adopters of Google Cloud's Gemini Enterprise for Financial Services, for agentic Financial Research workflows. Now previewing for corporate banking and capital markets, the platform runs research end to end with confidence scores and source citations.
Citi Wealth, Lloyds, Macquarie, BNY — Citi Wealth, Lloyds Banking Group, Macquarie Bank and BNY are among financial institutions adopting Google Cloud's Gemini Enterprise for Financial Services, giving staff agentic research workflows. Its Google-managed Financial Research agent runs research end to end on 50-plus specialised financial skills.
HSBC, Standard Chartered, Barclays, Citigroup, Deutsche Bank — HSBC, Standard Chartered, Barclays, Citi and Deutsche Bank are among global banks adopting Ant International's Falcon Time-Series Transformer (TST) 2.0 for AI FX forecasting and treasury hedging. Ant International's Kelvin Li says the model forecasts cash-flow and FX-exposure positions with over 90% accuracy and can lower hedging and allocation costs over 60%. None is independently evaluated.
Experian — Experian has added personalised credit scores to its ChatGPT app, calling it a UK first. Inside ChatGPT, users can now see their Experian credit score, its history and a factor-by-factor breakdown.
Absa — Absa's agentic AI customer service chatbot, which understands natural-language queries and asks follow-up questions, has topped 1.6 million users and handles over 100,000 queries a month. Group chief information and technology officer Johnson Idesoh reported the agentic assistant's figures.
DBS Bank — DBS Bank is scaling agentic AI in corporate banking to about 1,500 relationship and credit-risk managers globally after a 150-person pilot, with specialised agents covering more than 70 tasks and drafting review-ready credit memos. Credit memos and related work can take up to 40% of these bankers' time, which DBS aims to cut by at least 30%, says group head of institutional banking Han Kwee Juan.
Sun Life Canada — Sun Life's financial advisers have two AI tools: Notes Assistant, which summarises client meetings, drafts follow-up emails and compiles action items, and a new AI-powered adviser concierge answering product, policy and process questions. The insurer says Notes Assistant saves 15 to 30 minutes per client session and the concierge has handled over 11,000 client conversations since launch. Cutting admin frees advisers for client relationships, said Jessica Tan, executive vice-president and president of Sun Life Canada; Notes Assistant won a 2026 CIO Award Canada.
National Bank of Serbia — Retailers must now publish daily price lists, feeding National Bank of Serbia (NBS) AI price tracking that produces short-term inflation estimates, says Governor Jorgovanka Tabakovic. The goal: an artificial intelligence model for inflation estimates showing price direction before official statistics arrive. NBS calls it a supplement to the Statistical Office's monthly reports, not a replacement.
ruya — UAE digital bank ruya has started deploying agentic AI in banking operations with partner Magure. The goal is to become an AI-native bank.
Airtel Money Uganda — Fraud detection in financial services now relies on machine learning at Airtel Money Uganda, which flags risky wallets for human checks rather than freezing them outright. A robotic process automation tool reconciles balances, and staff view product results through a live dashboard. This AI-driven fraud monitoring covers reconciliation and analytics alike.
ABN Amro — ABN Amro's second-quarter 2026 results list nearly 50 AI use cases in production at the Dutch bank. Among them are GitHub Copilot for developers, a knowledge assistant helping analysts with anti-money-laundering and know-your-customer work, and the bank's first customer-facing generative-AI voice-bot.
FINCA Bank — FINCA Bank has deployed Kyrgyzstan's first Kyrgyz-language voice AI assistant for its contact centre. Built in-house with proprietary recognition and synthesis models, it handles inbound and outbound calls plus live chat, letting people talk naturally rather than pick through menus. A banking voice bot transfer to a live agent stays available on request.
Interbank — Interbank has become the first Peruvian bank with a direct Anthropic agreement, deploying Claude Enterprise to an initial group of staff including knowledge workers and specialist technology teams. Training and support programmes accompany this enterprise AI in banking rollout to keep adoption responsible, the bank says. Claudio Querol, vice-president for data, analytics and AI, plans to extend the tool gradually to more staff and processes.
Bir — Bir has dropped a chatbot named Ey-ay into the m10 digital wallet, the first AI payments assistant fielded in Azerbaijan and a product of the Caucasus region's earliest joined-up e-commerce, banking and payments ecosystem. Everyday transfers and the rest of the menu now run as conversational payments through plain chat, riding a swift national retreat from cash. The group counts over 5 million customers, tied in with Trendyol Azerbaijan and BakıKart.
SEB banka — SEB banka, the Swedish SEB Group's Latvian subsidiary, runs an AI voicebot banking service that lets customers lock a card or halt internet banking access on their own, without a human consultant. Launched to cover urgent requests, the system answers calls between 8pm and 8am on weekdays and non-stop across weekends. This voice assistant for banking targets time-sensitive emergencies rather than routine enquiries.
Absa Bank Kenya — Credit risk cycle times at Absa Bank Kenya dropped 80 to 90 percent, hours standing in for weeks, since SAS Viya launched atop Amazon Web Services. A new Centre of Excellence ships models 50 percent sooner under this AI credit risk management drive, phasing out scripts once spread across 500-plus tools. Retail lending signals a pivot toward AI-driven credit risk modelling.
NCBA Group — PropwizAI, the assistant embedded in NCBA Group's new PropertyDuka platform, studies a buyer's income and tastes to suggest fitting houses, mortgages and repayment routes. Billed as East Africa's first AI-powered property marketplace, PropertyDuka merges listings, construction loans, insurance and legal aid into a single online hub, an AI-driven property platform spanning the whole home-buying journey. NCBA's wider digital marketplace has already signed up almost seven million users and moved about KSh4 billion in transactions.
Absa Bank — Absa Bank's credit risk management AI slashed model reporting timelines by 80 to 90 per cent, compressing weeks into hours, since SAS Viya went live on Amazon Web Services. Fresh models now enter service 50 per cent quicker, spanning upwards of 500 models within Absa's retail lending portfolio, while live dashboards and built-in guidance sharpen credit risk assessment in banking.
CIBC — CIBC has introduced CAI 2.0, an agentic-AI workspace where employees hand complex, multi-step tasks to AI agents that plan, coordinate and finish them. The bank calls it the first enterprise-wide platform of its kind in Canadian banking, and says predecessor CAI reached over 50,000 employees worldwide, with about 20,000 daily users.
National Bank of Egypt (NBE) — National Bank of Egypt (NBE) took three Digital CX Awards 2026, recognising AI in banking used throughout its core operations. The NBE customer engagement engine weighs each customer's product holdings, demographic data and transaction behaviour to suggest their next best action. Other banking AI tools predict branch cash needs to support liquidity management, and speed document verification so approvals rely on fewer manual reviews.
I&M Bank Rwanda — Pindo built the Kinyarwanda Voice AI Banking Agent for I&M Bank Rwanda, the country's first such tool, so customers now bank purely by talking, in Kinyarwanda or English, over an ordinary phone call, no app or data connection needed. This voice banking service launched under the 'Tsinda na Banki Muhorana' campaign.
Itaú Unibanco — Itau Unibanco has released 'ia.i', an in-app AI banking assistant, to 300,000 clients, who can do their banking and seek financial guidance in natural language, spoken or typed. Built on a multi-model architecture, it should reach the whole client base by December.
Access Bank Plc — Access Bank Plc's Access SME App gives Nigeria's micro, small and medium enterprises an AI-powered SME management app with embedded funding. Record-keeping, stock oversight and staff supervision run unattended, feeding owners an AI-generated business insights tool, launched at its MSME Digital Growth Conference.
Bank of Maldives — Bank of Maldives customers can now file banking service requests directly through Aaya, an example of an AI virtual assistant in banking absorbing tasks once handled by a separate messaging feature. The AI-powered digital assistant, built into the mobile app and Internet Banking platform, answers in real time in Dhivehi and English.
Banco Popular Dominicano — Banco Popular Dominicano now uses AI virtual assistants to arrange bank-card delivery, calling customers to confirm addresses and chase new cards, renewals and replacements. Customers may switch to a human agent at any time, and the bank frames the AI as support for its staff.
ZamZam Bank — ZamZam Bank and the Addis Ababa Chamber of Commerce now back small firms lacking pledgeable assets, using AI-based SME lending to judge applicants purely on repayment capacity. Loans of 60,000 to 1 million birr flow to chamber members via Ansar Digital Financing, built with Kifiya Financial Technology. This algorithm-driven credit assessment had already reached over 10,000 SMEs with more than 1.1 billion birr beforehand.
Stanbic Bank Ghana — Stanbic Bank Ghana is countering financial crime through AI fraud detection in banking, deploying tools that track spending habits and flag irregularities before damage spreads. Felicity Osafo Sampong said the technology reads shifting customer patterns to catch suspicious activity as it happens, replacing after-the-fact review with real-time transaction monitoring powered by AI.
Zurich Insurance Japan — Zurich Insurance Japan runs 12 AI agents at once for new-contract review — medical, environmental, moral risk and fraud checks — marking an early case of AI underwriting insurance locally. New Business 2.0 targets 60 percent automatic assessment and a 77 percent cut to manual processing time. Zurich calls this AI-driven underwriting deployment the first multi-agent rollout across a full underwriting workflow at a Japanese life insurer.
First City Monument Bank (FCMB) — First City Monument Bank has built AI chatbot in banking functionality into its app, naming the tool Temi. The bank paired this with reward points, mutual fund access and instant virtual cards, sharpening its digital-first edge in Nigeria's crowded online banking market; Temi acts as an AI-powered banking assistant for routine support.
Coronation Insurance Plc — Nigeria's Coronation Insurance Plc now offers AI vehicle inspection on its digital onboarding portal, where motor insurance buyers photograph their cars rather than awaiting a manual check. This photo-based inspection relies on image recognition technology, speeding underwriting and claims assessments.
National Bank of Georgia — The National Bank of Georgia's NBG AI Platform, an AI in central banking project, is a Top 12 finalist in the 2026 Qorus-Infosys Finacle Banking Innovation Awards (AI-led Value Creation). Built in-house without external vendors, this central bank AI gives employees AI agents plus specialised chatbots for analytical and supervisory work.
TBC Bank — TBC Bank, Georgia's largest bank, gives 600 employees direct access to company data through Genie, an AI-powered data analytics tool built on Databricks, returning answers in seconds instead of via disconnected on-premises SQL servers. Work once taking a month now moves far faster, reports shrank from 14 weeks to two days, and the analyst team fell by 50%. Most headquarters staff are expected to use this AI in banking tool daily.
Nedbank — Nedbank's Quick Loans, short-term lending in the Money App, takes credit decisions from fintech Jumo's engine for AI credit scoring. Affordability checks run live on alternative financial data, with loans of R500 to R50,000 paid in as few as five minutes, for clients with thin or no credit bureau history too.
Bank ABB — Bank ABB threw open AI-nur and AI-khan to the whole client base once trials wrapped, billing the duo as among the earliest voice-driven AI banking assistants in Azerbaijan and the wider region. Conversational banking has drawn 230k people and 2.5 million exchanges. The lender keeps 180 generative and predictive models in production, worth above 35 million AZN in positive impact on the bank's own count.
TD Bank Group — Canada's TD Bank Group has cut mortgage pre-adjudication times from 15 hours to under three minutes using an agentic AI model built by its Layer 6 unit. This agentic AI in mortgage underwriting also handles Home Equity Lines of Credit applications, drafting summary memos after joint work across TD's Global Technology & Solutions, data, Real Estate & Secured Lending and risk teams. The underwriting automation gain is an early internal estimate, unconfirmed by outside review.
Zurich Insurance — Zurich Insurance runs agentic AI underwriting in five countries, including Switzerland, after a 90-day rollout with Cytora. Its AI risk digitisation platform converts unstructured, multilingual submissions into structured risk data, cutting triage from 75 minutes to 15 and lifting straight-through processing to 95%. Zurich plans expansion to over 20 markets within 16 months.
Heirs Insurance Group — Heirs Insurance Group has introduced Prince AI, an AI insurance assistant answering queries in languages such as English, Yoruba, Igbo, Hausa, French and Chinese. On WhatsApp, the SimpleLife mobile app and the insurer's site, the generative AI handles policy purchases, renewals, and filing and tracking claims.
Banque Mauritanienne pour le Commerce International (BMCI) — Banque Mauritanienne pour le Commerce International now answers customers around the clock through Aly, an AI virtual assistant banking service on WhatsApp staffed 24 hours a day, seven days a week. It routes account queries, points users to the right department and speeds up access to banking information. BMCI claims a first among Mauritanian lenders, becoming the earliest to weave a smart conversational agent through customer service.
Sumitomo Mitsui Financial Group — Sumitomo Mitsui Financial Group and Sakana AI built an AI proposal generation tool for corporate banking at Sumitomo Mitsui Bank. Agents split the work: gathering facts, analysing, framing hypotheses, verifying results, in one wholesale proposal drafting pipeline. SMBC's figures show weeks cut to minutes or hours.
AYA SOMPO Insurance — iVA, an Insurance Virtual Assistant from AYA SOMPO Insurance, is Myanmar's first AI chatbot insurance tool built into a mobile app, the firm says. Housed in the My AYASOMPO app, it answers product and policy questions in Myanmar and English with 24/7 real-time replies, no human needed. This AI-powered insurance assistant is meant to help partners and agents meet customer needs faster.
Capitec Bank — Capitec Bank has equipped almost 5,000 employees with AI tools and runs agentic AI in credit processing for business banking, Capitec's 2026 integrated annual report says. Capitec Pulse AI, built in-house, feeds client support agents live client information, and generative AI now works in production for compliance. The bank says the rollout is well past the experimental stage.
MobileMoney Fintech Limited — MobileMoney Fintech Limited has built AI fraud detection in mobile payments into Ghana's top mobile money service. Chief Executive Shaibu Haruna said the tool tracks spending patterns continuously, retracing payment paths to flag anomalies that static, rule-based checks miss. That signals a move toward AI-driven transaction monitoring platform-wide.
CIMB Niaga — CIMB Niaga has deployed AI agents for relationship managers, starting with a Relationship Manager AI Agent built with Google Cloud and Artefact as a strategic thought partner for bankers. A Contact Center AI Agent now also partners service teams live during customer calls; the bank's AI Center of Excellence developed both.
Scotiabank — Scotiabank has launched Scotia Intelligence, a unified enterprise platform that gives employees governed AI tools and the Scotia Navigator assistant for everyday work, custom assistants and AI-assisted coding. Branches and contact centres were early adopters of purpose-built tools; the bank says AI now handles over 40 per cent of contact-centre client queries and processes and routes about 90 per cent of commercial banking emails.
Daido Life Insurance — Generative AI in benefit claims assessment reaches full deployment at Daido Life Insurance this April. Its Benefit Payment Assessment AI, built on RAG-based retrieval, handles medical judgment on hospitalisation payouts, replacing in-house doctor review in around 70 percent of cases and saving roughly 1800 working hours yearly. The insurer calls this AI-driven medical judgment for insurance claims a sector first.
Joyo Bank — Joyo Bank's banking AI deployment, JOYO AI AGENT, runs on a closed cloud network cut off from the internet, offering translation and document masking. This AI agent for banks also trials loan approval review and automatic business document drafting. Athena Technologies co-developed it, building on a proof of concept from July 2025.
Starling Bank — Starling Bank customers can now handle money tasks through Starling Assistant, billed as the UK's first agentic AI financial assistant for current account holders. Voice or natural language prompts let users set savings goals, arrange bill payments and get insights via this AI-powered banking assistant.
NLB (Nova Ljubljanska Banka) — NLB, Slovenia's largest banking group, built an enterprise Agentic AI platform with Adastra, launching its first AI use case within five months. NLB applies AI in banking to high-volume documents, credit and compliance fact-gathering, and employee decision-making. Centralised governance, role-based access and cost controls cover bank-wide AI agents.
SanlamAllianz General Insurance Ltd — Customers of SanlamAllianz Mauritius can now message MIRA (My Insurance Robot Assistant) on WhatsApp for 24/7 help in English or French. Built as an AI insurance assistant, the bot fields common policy queries and steps in when staff are unavailable outside office hours, giving round-the-clock support via an AI conversational assistant for insurance queries. SanlamAllianz paired the debut with Zinsure, a fresh digital-first insurance product line, on 5 March 2026.
Britam General Insurance — Britam General Insurance settles light motor claims in two hours, against the industry's five working days, through AI claims assessment at Britam Centre's Nairobi drive-through. AI grades damage photos in about 15 minutes, an electronic claim form reaches the owner's handset and in-house checks take 30 minutes. Comprehensive motor policyholders qualify for AI-assessed claims.
Sumitomo SBI Net Bank — Sumitomo SBI Net Bank is beta testing NEOBANK ai, an AI banking agent built on generative UI. The assistant reads users' plain-language goals and renders screens live for fund transfers, procedure navigation and debit card spending analysis. The lender calls itself Japan's first bank with an AI agent for banking that builds interfaces per request.
Cova — Cova switched on AI motor insurance pricing for compulsory cover in Cameroon on 26 February 2026: film the vehicle's carte grise, or a lodging receipt, and the regulated tariff returns in one click, renewal premiums prepayable remotely. Reading fiscal horsepower, use and energy type off the document against the standing tariff grid cuts keying mistakes and compresses sign-up times, by Cova's own account. Collection runs through mobile payment aggregators Maviance, UFIPAY and MaMoni Pay, plus Express Union Microfinance and MaMoni Finances counters. Managing director Virginie Pouna Ngomi framed the automated premium quotation as cover simpler and more dependable, licensed brokers and partner insurers undisplaced; Cova holds a broker licence in the CIMA zone and came via the BimaLab Africa accelerator.
Sumitomo Mitsui Banking Corporation — Sumitomo Mitsui Banking Corporation started SMBC AI Operator, a generative AI call centre line with technology from the Japan Research Institute and IBM Japan, on 25 February 2026. 24/7 it takes free-speech calls on the Olive personal finance product's application steps, campaigns, annual fees and service content, answering from bank reference data and passing callers to a human operator for identity checks during call-centre hours. The voice operator adapts phrasing and tone to each caller's way of speaking.
Czech National Bank — The Czech National Bank has granted the first six permits under the European MiCA regulation, out of 248 filings, leaning on AI in crypto-asset licensing for the paperwork. Software trialled in the bank's ČNB Lab innovation unit runs on hardware the bank owns; the machine review of licence files flags gaps, spots contradictions across papers and audits content. Jan Procházka stressed that granting or refusing a permit always rests with ČNB experts, technology acting only in support.
Raiffeisenbank — Raiffeisenbank staff can call on about 130 varieties of AI assistants for bank employees, chief executive Igor Vida said. An auditor part-way through an inspection, by the bank's own example, hands the findings over and gets back a finished report. Vida cast the workplace AI rollout as lifting output per head instead of shrinking headcount, backed by in-house teaching on digitalisation and these tools.
ANZ — ANZ has put Salesforce's Agentforce 360, AI agents for business bankers, into its bankers' new customer relationship management tool, claiming to be the first Asia-Pacific bank to deploy it at scale. The nationally rolled-out, Agentforce-backed CRM replaces a legacy platform and merges data from 20 separate systems into one dashboard, cutting time bankers spend switching systems, ANZ says.
Fonds de Garantie de l'Entrepreneuriat au Congo — The Fonds de Garantie de l'Entrepreneuriat au Congo opened Bokeli, an AI business-plan platform, at Kinshasa on 3 February 2026, aimed at micro, small and medium firms plus artisans lacking the security lenders usually want. Bokeli guides a founder through a bankable document with situational prompts and samples, inserts projections at a click, saves to PDF or Word, then pushes the file into the fund, where applicants watch its progress themselves until cash lands. Operations chief Ibrahim Ucircan said most submissions arrived in poor shape whatever province they came from. Director General Laurent Munzemba Kompa said the automated plan builder kills off gripes about sluggish handling.
Manulife — Manulife has launched a redesigned e-application and upgraded MAUDE, its proprietary AI underwriting engine, that together auto-approve qualified life insurance applicants in as little as two minutes. Cases the engine cannot approve go to a human underwriter. By December, the chief underwriter said, 58 per cent of eligible cases were auto-approved through MAUDE, up 56 per cent on pre-launch levels.
AXA Mansard Insurance — Snap a photo, skip the inspector: AXA Mansard Insurance now runs AI digital vehicle inspection in Nigeria, built alongside Curacel, covering new comprehensive policies, renewals and claims. Machine-driven photo review has taken over from on-site assessors.
MTN Ghana — MTN Ghana built AI fraud detection mobile money algorithms that intercept fake MoMo alerts before phones receive them. Abdul-Majeed Rufai said the system flags impersonation even when scammers alter headers or misspell words — part of AI-driven fraud detection in mobile money services spreading across telecom operators.
Visa — Visa is piloting an AI financial assistant with US banks that shows cardholders in-app spending insights, fields natural-language questions on their activity, and lets them lock cards or set alerts mid-conversation. Visa says the pilot starts in August 2026, going global after the US launch.
Commonwealth Bank of Australia's agentic AI fraud detection, built in-house in about three months, helped cut fraud losses more than 20% in the first half of its 2026 financial year year-on-year, the bank says. Watching over 80 million signals a day, the fraud agent helped create or update three-quarters of card-fraud rules, with human approval. James Roberts, general manager of fraud and scams, reported this in April 2026; no independent evaluation is published.
The UK Financial Conduct Authority's AI Live Testing is a voluntary supervisory service letting cohorts of about five to ten firms trial live AI proofs-of-concept with the regulator. Part of the FCA's AI Lab, it began in October 2025, with a second cohort from April 2026.
Tokio Marine & Nichido Fire Insurance — Tokio Marine & Nichido Fire Insurance put AI contact centre tools across its customer and agency lines on a PKSHA Technology and CTC platform. PKSHA VoiceAgent captures callers' identity and purpose from natural speech before routing, Speech Insight provides live transcription and suggested answers for operators, and Knowledge Stream drafts knowledge-base entries afterwards. It expects up to about 30 per cent shorter handling on customer calls and about 10 per cent on agency calls, freeing operators for specialist work.
Bank of Korea — The Bank of Korea runs BOKI, an in-house AI system, on its internal business network, where staff dig through the bank's document archive, even reading tables and graphs from reports. Data never leaves the building, a design built with Naver Cloud after a late-2023 tie-up, reportedly making it the world's first central bank to build and run internal AI this way.
Banco BPI — Portugal's Banco BPI has put the Harvey legal AI platform across its whole legal department after a pilot the bank says raised productivity in the team's core work. BPI says the platform frees its lawyers' time from procedural tasks, and head of legal Luís Graça Moura says practical gains are already showing.
Uniqa Biztosító Zrt. — Uniqa Biztosító Zrt. has widened NiQA, its AI claims handling platform, from burst pipes to glass damage, and artificial intelligence now touches one household claim in three in Hungary. Thorniest category first, with every notification the eventual ambition. Automated claims assessment has paid out inside 24 hours since mid-April, and Uniqa's own forecast puts most retail claims automated by 2028.
Conister Bank (Manx Financial Group) — Conister Bank, part of Manx Financial Group, now applies AI in banking redress work, automating its Discretionary Commission Arrangements restitution process for far less than a third-party processor would charge. The tooling sorts through bulk complaint correspondence, pulls out key data points and drafts customer replies, though humans still sign off every final decision. This automated redress handling forms part of the wider AI strategy the Isle of Man-based group set out.
CMB Monaco — Monaco's CMB tops Europe in generative AI wealth management, merging its systems onto Avaloq and BlackRock Aladdin Wealth via their June 2023 alliance. Advisers now receive 'Auto Commentary', translating dense portfolio figures into brief written notes, while back-office corporate actions run through AI-led automation.
SBS Bank Eswatini — SBS Bank Eswatini runs treasury work on Intellect Design Arena's eMACH.ai Treasury platform, an AI-powered banking treasury tool covering money markets, foreign exchange and fixed-income securities at 100% automation. Automated SWIFT messaging cut manual effort by 90%. The move backs growth from building society into commercial bank, after a provisional licence from the Central Bank of Eswatini.
One Zero Bank — One Zero Bank now lets customers pull their own financial data through ChatGPT and Claude, skipping the lender's app and website entirely. This puts the Egyptian bank among MENA's earliest adopters of agentic banking, where an external AI agent banking integration retrieves account details directly when prompted through a conversational tool the customer already uses. One Zero calls it a step toward financial infrastructure for the AI agent era.
Commercial Bank of Ethiopia (CBE) — Commercial Bank of Ethiopia now runs AI-driven fraud detection in banking, scoring each payment for risk via an AI-powered fraud management system built on behavioural analytics. A flagged score triggers a freeze, decline or extra check, spotting hijacked accounts, mule networks and coordinated fraud rings. Models keep updating as tactics evolve.
Wema Bank — Wema Bank embedded SAW, an AI voice banking assistant, within ALAT: The Evolution. Spoken requests let customers check balances, transfer cash and settle bills across English, Yoruba, Igbo or Hausa. Facial recognition alongside voice-controlled financial services checks confirms who is speaking.
King Price Insurance — Half the wait has vanished from routine personal lines claims at King Price Insurance since a new engine took over deciding how each case gets handled, sending turnaround times down 50 per cent. This early instance of AI claims processing in South African insurance sits inside a broader enterprise AI division backing every claim, built around the AI-driven claims risk engine. Overhauled optical character recognition and purpose-built fraud detection models round out the shift, catching syndicate behaviour nobody had spotted before.
Mizuho Financial Group — Mizuho Financial Group lets non-engineers build AI agents on a new Dify Enterprise AI agent development platform, phased in from May 2026. A corporate banking trial cut task time 41.8 percent on average, 52.2 percent for junior staff; comment generation accuracy in AI in corporate banking hit just 6.3 out of 10.
MS and AD Insurance Group — AI image verification in insurance claims now runs at Mitsui Sumitomo Marine and Aioi Nissay Dowa, under MS&AD Insurance Group, catching fabricated or altered damage photos and receipts. Tech Union helped build it, marking Japan's first such AI fraud detection for claims documents among non-life insurers, curbing increasingly varied fraud.
Aomori Michinoku Bank — Every branch of Aomori Michinoku Bank now runs neoAI Chat, five months after a March pilot began. Staff use this AI agent platform for banking tasks: drafting, summarising, checking internal rules, and gathering information. The rollout reflects momentum behind agentic AI in financial institutions across Japan.
Gunma Bank — Gunma Bank's FutureBANK AI HUB, a generative AI agent platform built with Future Architect, guides staff through lending operations using internal manuals and historical loan records. This is AI-assisted lending in practice: within a fortnight of launch, roughly 60 percent of eligible employees had tried the tool, logging close to 4000 uses. The uptake points to strong early adoption of AI in loan processing at the bank.
Aiyl Bank (ABank) — Aiyl Bank's ABAI chatbot and AI document analysis in banking trimmed manual workload at the Kyrgyz lender in 2025, freeing 9 thousand 600 working hours. Regulatory paperwork now moves through the AI assistant instead of staff, while ABAI fields client queries. This AI-driven automation in banking shifted staff onto harder, client-focused work.
Bank SinoPac (Yongfeng Bank) — Yongfeng Bank rolled out "Licai Copilot", an AI-driven wealth management tool built on Microsoft Azure AI, aimed at Taiwan's high-end wealth management clients, a segment the bank values at NT$2 trillion in assets. Advisory staff gain four capabilities through the system: rapid client lookup, an internal intelligence feed pulling nearly 500 news sources, quick client profile writeups, and simulated compliance drills replacing standard training sessions. The rollout signals the bank's broader shift toward AI wealth management tool adoption across its advisory teams.
De Surinaamsche Bank (DSB) — De Surinaamsche Bank (DSB) rolled out DSB Buddy, an AI banking chatbot answering client queries on offerings through WhatsApp at any hour. The Suriname lender's virtual banking assistant sits within a modernisation push with an entirely online personal loan, freshly introduced POS terminals and corporate prepaid cards.
Bupa Hong Kong — Faster claims turnaround, stronger fraud checks and smoother regulatory compliance are the targets of a fresh five-year Business-Process-as-a-Service pact between Cognizant and Bupa Hong Kong, built around AI in health insurance claims processing. AI-driven claims automation sits at the core of the arrangement, Cognizant's first such BPaaS launch in the region, meant also to lift customer satisfaction.
MBANK — MBANK's AI customer service assistant, AI MBANK, settles 79 per cent of Kyrgyzstan queries unaided. This virtual banking agent, unveiled at Bishkek's Mtalx event, fielded over 2.7 million chats in Kyrgyz, Russian and mixed languages in 2025, via the MBANK app, WhatsApp and Telegram. Figures are unaudited.
Mapfre Perú — Under one second: that is now the approval time for 60% of AI vehicle inspections at Mapfre Peru, down from a wait of up to five days under the old manual process. Computer vision and automated decision models inside the insurer's Control Expert platform drive the instant automated vehicle inspection approval. Mapfre credits the overhaul with reshaping its claims workflow and boosting customer satisfaction, though these are the company's own figures.
bunq — bunq's Finn, an AI chatbot for banking customer service, resolves roughly 97% of support interactions, replying within 47 seconds on average across 38 languages at 90% satisfaction. This AI-powered virtual banking assistant also covers spending insights and travel tips for the Dutch neobank, escalating harder cases to staff.
Gebäudeversicherung Bern (GVB) — Berna, an AI voicebot for claims intake, now fields the flood of calls Gebäudeversicherung Bern (GVB) receives after major loss events, built with Spitch and Adnovum. It copes with Swiss German, manages multiple conversations at once, and converts spoken loss reports into structured dossiers ready for staff. Anyone on the line can still request a human specialist, part of GVB's push into AI-powered insurance claims handling.
Nexent Bank Romania — Nexent Bank's Romanian arm now runs AI credit decisioning through SAS Viya, turning loan approvals into a minutes-long process rather than days. The platform clears 70% of credit card limit requests without staff review, drawing on machine learning for behavioural analysis and sharper credit scoring. Agile methods let the bank, part of the Netherlands-based group, finish this initial AI-driven lending build in six months.
Rizal Commercial Banking Corporation — Rizal Commercial Banking Corp engineered an AI credit decisioning engine now driving consumer loan approvals bank-wide. Reviewers face lighter workloads, favoured customers get sign-off within a day, and alternative data brings thin-file applicants into reach. This automated loan approval push has slashed processing times bank-wide.
The Company for Cooperative Insurance (Tawuniya) — The Company for Cooperative Insurance (Tawuniya), Saudi Arabia's largest insurer, has deployed an AI medical underwriting system to shrink claims leakage and tighten premium pricing accuracy. Anonymised patient histories, treatments and hospital bills feed the tool, flagging overcharges, duplicate claims and fraud. Lotte Data Communication (South Korea) and Synapxe (Singapore) co-developed the AI-based insurance underwriting technology, now live across employer health plans for over five million members, with hospitals already onboarded in Riyadh, Jeddah and Dammam.
Trinidad and Tobago Stock Exchange (TTSE) — TOBI (TTSE Online Bot Interface) covers trading activity, listings, corporate announcements, dividends and settlement procedures at the Trinidad and Tobago Stock Exchange, round-the-clock. This AI chatbot for investors, a digital investor assistant, serves newcomers, brokers and listed firms on the exchange's website. It backs a wider digitalisation drive for capital-market infrastructure.
Islami Bank Bangladesh PLC — Islami Bank Bangladesh PLC debuts CellFin Echo inside CellFin, the nation's first tool for AI banking in Bangladesh driven purely by speech. Speaking plain English, users check balances, move funds, settle bills, pull ATM codes or view statements via this spoken assistant. CellFin already counts nearly 5.7 million users.
Union Bank of the Philippines — UnionBank is the Philippines' first bank, and among ASEAN's earliest, running Amazon Quick Suite for natural language data queries in banking. Over 200 staff in operations, technology, product and customer-facing roles skip IT, pulling insights via plain-language data querying. UnionBank expects 3 to 5 times faster request handling.
Alborz Insurance — Alborz Insurance (بیمه البرز) has rolled out insurance claims automation in Iran through a new assistant unveiled at the International Insurance and Development Conference, letting policyholder paperwork clear without any staff touching it. Files uploaded to the system get checked in under two minutes, then pass straight through to payout. The insurer bills it as the industry's first AI-powered claims handling tool, aimed at cutting errors and quickening settlements.
Igloo — Igloo applies AI in insurance across the Philippines, spanning product design, sales and claims. Drop-off fell 24.7% once a recommendation tool steered buyers to suitable plans, 70% of whom accepted it, while a digital claims helper lifted accuracy to 98% and let policyholders file instantly. Agents get round-the-clock quoting and payments via chatbot.
BEC member banks — Salesforce's Agentforce platform now powers AI agents in banking advisory services across BEC's member banks, the shared IT centre for Danish banks. Advisors already lean on a live generative tool to draft customer replies inside online banking, while a second batch of agents, still piloted, works as an on-demand assistant, condensing messages and rendering them in English. This deployment of AI agents for bank advisors covers routine, command-driven tasks rather than open-ended judgement calls.
LAQO / Croatia osiguranje — LAQO, Croatia osiguranje's digital brand, is Europe's first insurer selling travel insurance directly through an AI agent, on WhatsApp. Pavle Pauq, the AI insurance sales assistant, queries trip details, offers extra cover, sets the premium and issues personalised payment links. AI-led selling followed LAQO GPT, which resolved 30 per cent of enquiries unaided.
Prime Bank PLC — Prime Bank PLC now offers Bangladesh's first AI-powered digital letter of credit management platform, Prime Banijjo. Clients get automated drafting, compliance screening and live tracking, easing paperwork and speeding turnaround across the LC lifecycle. First for AI in trade finance locally, per the bank.
SMBC Indonesia (Jenius) — Jenius, SMBC Indonesia's digital platform, folds AI in digital banking into signup, AI-powered lending decisions and fraud surveillance. Biometric checks confirm new users, an automated engine assesses loans instantly, and device intelligence spots suspicious behaviour, cutting unauthorised-transaction losses over 80%. A chatbot fields 70% of monthly enquiries too.
Banco de Guatemala (Banguat) — AVI, an AI virtual assistant economic data Guatemala tool from Banguat, sits inside Indicators GT, resolving plain-language queries across over 12 million financial records. It sketches charts instantly, stays live 24 hours daily, and toggles Spanish and English on any device — an AI-powered data-query tool for the public.
KakaoBank — South Korea's largest internet-only bank switched on AI money transfer banking on 24 November 2025, letting a spoken phrase like "Mom, 300,000 won" replace manually keyed recipient details. KakaoBank's app accepts that instruction by voice or text, with speech recognition active on Apple hardware — a first among domestic financial firms.
Groupama — AI in insurance operations at Groupama, one of Romania's largest insurers, cut CASCO quote turnaround by 70%, with 97% of eligible offers now handled on a single Gemini Enterprise platform from Google Cloud. Response times to policyholders fell 40% and paperwork checks dropped 50% within the tool's first month live. At AloGroupama, the company's phone support line, automated voice handling driven by AI-based customer service now closes out requests for over 1,000 callers each month without staff involvement.
Citi Kenya — Citi extended generative AI in banking to over 200 Kenya staff, running Gemini, ChatGPT and Grok inside gated frameworks tied to internal data. Automated summaries, draft reports and analytical support now replace manual routine tasks. Kenya ranks among Citi's first Sub-Saharan African markets for this AI tools for bank employees deployment.
CNA Insurance Canada — Select business lines at CNA Insurance Canada now run on mea Platform's Intelligent Underwriting AI, part of a November 2025 rollout the insurer plans to extend further, marking a fresh step in AI insurance underwriting. Structured and unstructured records feed the ingestion engine, which pulls out the data points needed for new business submissions without manual sorting. CNA credits the launch with fast turnaround and strong data accuracy, though that judgement comes from the company itself rather than an outside audit, underscoring the broader push toward AI-driven underwriting automation.
BPCE Group — Every one of BPCE Group's 100,000 staff can now access MAiA, its generative AI tool, with 50% firing it up daily for roughly 40 sessions apiece and the group logging over 2 million prompts a month overall. What started as a plain internal messaging assistant has grown into several dedicated bots tackling fraud prevention, legal services and customer relations, an example of AI in banking spreading through frontline roles. 75% of branch advisers already lean on these AI-powered tools routinely.
Bank Mellat — Bank Mellat's Dima app gives customers an AI banking assistant that carries out requests spoken or typed. On Android and the web, Dima also lets customers open accounts remotely, move money, request cards and use video banking. Bank Mellat claims Dima is Iran's first banking app whose voice-driven AI assistant adapts to users' age, gender and preferences.
Alliance Bank Malaysia — Alliance Bank Malaysia's Retail Loan Origination System (RLOS), built with CTOS Data Systems and JurisTech, has lifted application turnaround and staff efficiency by more than 30 percent, an early win for AI loan origination in Malaysian retail banking. The platform spans the whole lending journey, from lead capture and digital onboarding to decisioning, disbursement and credit administration, and mortgages, personal loans and credit cards now run on this automated loan approval system.
Rímac Seguros — Rímac Seguros now finishes vehicle insurance claims in one minute flat, down from a 48-hour manual slog, via AI claims processing. Peru's biggest insurer, over 30% market share, built a chatbot: drivers snap photos, share video, log the incident, and instantly pull coverage, deductible and repair shop info. This AI agent went live in November 2025.
FNB — NAVi, FNB's multi-agent virtual assistant inside the FNB App's Secure Chat, now handles fraud reporting, card replacement and debit order disputes, extending AI in retail banking beyond simple chatbots. Routine requests route automatically to NAVi, freeing advisors for harder cases; a pilot WhatsApp channel adds a verified line for banking queries. The phased South African launch underlines FNB's bet on an AI-powered banking assistant.
Banco de la Provincia de Buenos Aires — Argentina's Banco Provincia switched on PIA (Provincia Inteligencia Artificial) in November 2025, merging WhatsApp, website and social-media support into a single generative AI banking assistant for the Buenos Aires provincial lender. People asking about Cuenta DNI or Plazo Fijo accounts get tailored, conversational replies, while card controls, PIN resets and product lookups run non-stop with no staff needed. A once-fragmented helpdesk becomes one always-available AI virtual assistant for banking.
Nubank — Nubank's AI credit card limit increases in Brazil raised limits for some customers as its loan book grew 42% to US$30.4 billion in Q3 2025. Guilherme Lago said sharper credit analysis drove growth without matching risk. The AI-driven credit assessment also lifted revenue and profit.
Fiji Development Bank — Fiji Development Bank runs a 12-month AI credit assessment trial cutting loan turnaround for current customers. Its model blends records with behavioural signals, including WhatsApp and Viber history, yielding initial evaluations while FDB credit officers keep the final decision. Via the app, it spans unsecured loans to FJ$10,000 and secured loans to FJ$100,000, part of a shift toward AI loan approval for small businesses.
National Bank of Kazakhstan — An AI analyst for rate decisions is coming to Kazakhstan's central bank, offering independent input during Monetary Policy Committee sessions, the National Bank governor told the Central Asia Fintech Summit 2025. Built as a corporate AI tool, the system contributes views on agenda items without casting a vote, leaving ultimate authority with human officials — its role limited to automated economic analysis for monetary policy rather than decision-making itself.
Philippine Veterans Bank — PVB deployed an AI fraud detection platform for banking in 45 days flat, from signed contract to live production. Clari5 built the system, EFRM, an AI-powered Enterprise Fraud and Risk Management solution watching PVB's internet and mobile channels in real time. Its GenAI-ready design also helps the Philippine Veterans Bank meet Bangko Sentral ng Pilipinas compliance rules.
Česká spořitelna — A round-the-clock digital assistant now welcomes customers at Future Lab, the new Prague site launched by Česká spořitelna as its first nonstop branch. Staffed continuously, 24 hours a day, this AI avatar in retail banking works alongside human staff, offering guidance on financial wellbeing and handling routine service requests. Specialists trained in artificial intelligence collaborate there with traditional bankers, marking the lender's first use of this AI-powered virtual banking assistant across its branches.
Freedom Bank Kazakhstan — Freedom Bank in Kazakhstan now runs an AI voice assistant for entrepreneurs, answering 119 spoken commands in Kazakh and Russian. Owners settle taxes, shift funds, clear loans, run payroll and raise invoices aloud, hearing the CEO's voice reply. Voice-driven banking automation reaches small-business tasks.
DBS — Singapore's DBS has logged over 120,000 unique chats with DBS Joy, a generative AI assistant, since February 2025 trials. DBS Joy delivers AI in corporate banking 24/7 via IDEAL to corporate clients and SMEs, about 4,000 monthly, lifting customer satisfaction scores 23%. Corporate banking AI for Hong Kong and India is planned.
Intesa Sanpaolo — More than 3,500 employees at Intesa Sanpaolo have used SkillGym, an AI role-play platform for managerial assessment built on conversations with profiled avatars. Roughly 400 sessions finished in eight months, earning the bank a SMAU 2025 Innovation Award. The tool reflects growing use of AI-driven leadership evaluation through scripted dialogue.
LHV — Estonia's LHV counts nearly 1,000 employees on ChatGPT Enterprise, with over three-quarters embedding AI in banking into everyday workflows. A RAG tool steers workers through dense rulebooks and contracts; machine learning further powers marketing and software output. Founder Rain Lõhmus favours free-form testing over outright bans on new tools.
Freedom Insurance — Freedom Insurance runs AI fraud detection insurance checks in Kazakhstan via the Sergek intelligent monitoring platform. Road camera footage confirms reported collisions, while vehicle usage patterns feed risk scoring for voluntary auto policies. This AI-driven insurance fraud detection method builds accurate risk profiles for underwriting.
Türkiye Finans — Financially stretched customers of Türkiye Finans can now settle credit card and financing debts entirely on mobile, via the bank's new Digital Restructuring tool for AI-powered debt restructuring, skipping a branch visit altogether. Automated scoring reviews each case and drafts a bespoke repayment schedule, extending AI-driven debt restructuring into everyday consumer banking in Türkiye.
JS Bank — JSense arrives as Pakistan's first AI banking chatbot, launched by JS Bank within its mobile app. Round-the-clock coverage spans product help, transaction steps, branch details and account queries, powered by natural language processing — a form of conversational AI in banking.
UOB — UOB's AI branch bots in banking now cover more than 40 Singapore locations, giving counter staff quick answers on products and services. Built by UOB's Innovation Group on an in-house Build-Your-Own-Bots platform, they belong to a bank-wide set of eight bots spanning HR and branch operations. Over 20,000 of UOB's 32,000 staff have trained on Microsoft Copilot, part of its generative AI adoption.
Grupo Financiero Banorte — Banorte's Tesorería Inteligente: a free AI treasury platform mining millions of transaction records for payment trends, customer profiles and salary benchmarks. SMEs, corporations and government entities get access at no charge, making AI-driven cash management standard, not a paid extra.
Bede Bahrain — Bede Bahrain now offers AI customer service via two tools on Zain Bahrain's AI infrastructure: a WhatsApp/website generative AI chatbot and an AI-generated voice replacing IVR prompts. Callers hear naturally spoken, locally tuned guidance instead of scripted menus; chat users get human-like replies on products, services and application status. The aim: round-the-clock 24/7 customer engagement across both channels.
Mibanco — Mibanco, part of Grupo Credicorp and Peru's leading microfinance lender, has introduced AI conversational banking on WhatsApp, built with Amazon Web Services. Since launching by November 2025, the service has let customers open digital accounts, check balances and view product details in the chat; a February 2026 stage adds credit applications, QR payments and transfers. Mibanco aims for its generative AI banking channel to resolve 90% of queries, with only 10% passed to human agents.
Bahrain Islamic Bank (BisB) — Bahrain Islamic Bank now answers Sharia questions instantly via Al Murshid, an AI Sharia advisory platform trained on 2,000-plus Fatwas its Sharia Supervisory Board has issued since 1979. Unveiled at the AAOIFI-IsDB conference, it extends AI in Islamic finance compliance across daily banking. The Digital Banker Global Islamic Finance Awards 2025 named it Best AI Initiative.
Suncorp — Australia's Suncorp is rolling out agentic AI in insurance at full scale, starting with claims lodgement and customer service use cases. Suncorp's CIO expects AI agents to reshape customer experiences and end-to-end processes entirely. Home claims teams now use a smart product disclosure statement tool.
Tata AIA Life Insurance — Tata AIA Life Insurance in India now automates AI-powered insurance underwriting via its GenAI-powered Underwriting Rule Engine (URE), pairing machine learning and big data analytics so applications clear without human review. Medical history and lifestyle data feed the URE, which scores risk instantly and settles approval or rejection alone. Tele-medical exams gain AI-driven reflexive questioning, cutting reliance on physical checks.
Choco Up — Approving an SME loan now takes hours instead of days thanks to Choco Up's AI Credit Assessment Engine. Coding agents scan and validate figures pulled from bank records, sales trends and growth forecasts, enabling AI credit assessment SME lending for the Singapore-based growth financing platform - an AI-powered SME loan approval process reaching requests up to USD 200,000. The goal: closing Asia's US$2 trillion SME funding gap.
Starling Bank — Starling Bank now runs Scam Intelligence, an AI purchase scam detection feature in its app — a UK first. Built on Google's Gemini models, it examines photos of marketplace items or ads for red flags: inflated prices, faked pictures, pushy sellers, alerting customers before they pay. This fraud-screening tool works fast.
KBC Asset Management — Algorithms now steer portfolio choices at KBC Asset Management in Belgium, marking a shift in AI investment management from support tool to primary driver. Machine-learning systems trained over several years scan markets and companies at scale, generating data-driven insights the investment committee acts on directly rather than merely reviewing. Final accountability still sits with people applying artificial intelligence in fund management, not the software itself.
National Bank of Kuwait — National Bank of Kuwait extended Microsoft Copilot to its entire workforce, marking a full-scale bank-wide launch of the AI assistant. Staff now lean on it to draft documents, work through complex data sets, condense information into summaries, and automate repetitive tasks, a shift that reflects growing generative AI in banking. The achievement won NBK an AI excellence award at the Microsoft AI Summit in Kuwait, underscoring how far AI assistant deployment in banking has progressed within the institution.
Banco Nacional de Bolivia (BNB) — BNB (Banco Nacional de Bolivia) built AI in banking operations that freeze, release and suspend accounts under judicial orders with no manual steps, lowering mistakes and legal exposure. Felaban crowned its agentic AI tool, Orquestación de Agentes de IA, top AI Innovation for Banks entry at CLAB 2025, ahead of competitors tackling similar court-driven compliance work.
Fubon Bank (Hong Kong) — Hong Kong retail banking sees its first generative AI credit origination tool as Fubon Bank (Hong Kong) rolls out Moody's Automated Credit Memo for corporate lending. The system builds credit memos automatically, cutting the time spent gathering and reviewing data before origination decisions. Fubon is named as the first such lender in the territory to deploy this kind of AI-driven credit memo automation.
TD Bank Group — TD Bank Group's Layer 6 team built a GenAI virtual assistant for advisors: the Wealth Virtual Assistant, letting TD Wealth staff prompt for policy answers instead of searching multiple systems. This AI knowledge management system for wealth advisors frees time for personalised client advice.
CIBC — CIBC switched on CRTeX (CIBC Real-Time Experience) nationwide on 16 October 2025, an AI-powered personalisation engine for banking built in-house. Near real-time decisions merge with customer preferences so the tool shapes offers across phone, app, website and branch contact. This AI-driven client engagement platform, tied into CIBC's existing systems, keeps the experience uniform.
Mercado Pago — More than 60 financial tasks now run through spoken or typed requests, thanks to a new AI voice assistant for banking that Mercado Pago rolled out across Brazil. Balance enquiries, Pix transfers, savings scheduling and expense sorting all fall under this virtual assistant for finance, built on proprietary technology from Mercado Livre and reaching the customer base in stages. The launch marks the fintech's debut in this category anywhere across Latin America.
Singlife — Agentic AI insurance debuts at Singlife, a first for Singapore's insurance sector via new ties with Salesforce Agentforce. Materials in Salesforce Data Cloud—product manuals, training guides, FAQs—feed an AI agent customer service system that sharpens staff precision and pace on enquiries. Adviser representatives gain the same access afterward.
Bank of Singapore — Source-of-wealth report drafting drops from 10 days to roughly one hour at Bank of Singapore, OCBC's private banking unit, following its rollout of an agentic AI wealth management tool called Source of Wealth Assistant. Relationship managers upload client files, and the system pulls out, organises and combines the relevant details automatically inside controlled processes. A human still checks and signs off every report before it goes out, and the bank's approach reflects growing use of AI in wealth management across Singapore.
Aditya Birla Capital — Aditya Birla Capital added four generative AI tools to its D2C platform ABCD: Sales Assist, Service Assist, Audit Assist and Marketing Assist, launched at Global Fintech Fest 2025. Agentic AI for sales and service now covers onboarding, underwriting and claims management, backed by GenAI voice bots for tele-sales and queries — an AI-powered customer service automation push across India.
Intercontinental Trust Limited — Mauritius's Intercontinental Trust Limited now runs Amelia, an AI receptionist for financial services, reportedly the first such tool at a local management company. This agentic AI system for client queries fields calls and recurring questions on offerings and vacancies, logging exchanges securely and freeing staff for higher-value work.
Sancor Seguros — Sancor Seguros built Clara in-house: the Argentine market's first conversational AI voice agent for insurance. It rings early-stage debtors, converses in plain language, hooks into core systems, and fires off WhatsApp payment links. Already live for affiliates, the AI-powered voice assistant in insurance is billed by the firm as a sector milestone.
If Skadeforsikring — Denmark's largest insurer opened IfGPT, a generative AI chatbot for insurance customers, to over 200,000 Danish private customers, using if.dk material to handle home, car and travel cover queries. If expects this tool plus TopGPT will jointly resolve beyond 200,000 enquiries yearly, an AI-powered virtual assistant setup spanning both.
HNB Assurance PLC — HNB Assurance deployed HANA, an AI insurance chatbot for policy queries and claims. NLP and machine learning resolve complex requests and link to core systems, giving Sri Lankan customers 24/7 access via an always-on digital assistant.
STC Bank — STC Bank has deployed FOCAL, Mozn's AI fraud prevention platform, using real-time intelligence and automated compliance checks against financial crime. This supports AI fraud detection in banking, protecting customers as Saudi Arabia's digital sector grows. Mozn, a regional enterprise AI leader, calls FOCAL its flagship AI-driven financial crime prevention tool.
GFX Brokers — GFX Brokers built Kojo, an AI Eurobond sales agent running inside WhatsApp for Sub-Saharan clients. Built on Salesforce Agentforce with partner Evolf, it delivers live securities pricing, handles queries and lets buyers haggle bids and offers by text, cutting out calls and email. It turns the broker dealer's chat app into a bond trading channel.
tbi bank — Bulgaria now has its first AI agent in banking, deployed by tbi bank to reshape operational workflows. A task once needing 37 minutes of staff time now finishes in 2.76 seconds, as the tool pulls and refreshes records from public institutional databases without manual input. Running inside the SYNCRO AI platform alongside hundreds of RPA robots already in place, this addition to bank AI automation shifts employee focus toward higher-value strategic work.
Dhaka Bank PLC — Dhaka Bank PLC's eRin Device APP gives Bangladesh its first AI-driven digital loan platform, built with Samsung Electronics and Banglalink for financing smartphones and smart devices. AI-based lending now judges creditworthiness and speeds approvals for service professionals, micro-entrepreneurs and underserved communities.
Cathay Life Insurance (Guotai Life) — Taiwan's insurance regulator has approved Cathay Life Insurance to pilot facial recognition for policy signing, replacing 8 to 13 signatures per document with one signature plus two face scans. The face recognition insurance trial runs at two sites, finishing by end of 2025 before wider rollout of the biometric verification system.
Sense Bank — Sense Bank now runs an AI assistant for customer support built on the NovaTalks platform, pairing scripted chatbots for common enquiries with live operator backup, all linked to the bank's CRM. One system now covers contact centre calls, Viber and in-app chat, so routine messages get handled automatically rather than by staff. This omnichannel setup for AI in banking customer service pairs with a separate lending conveyor, where AI reviews documents and sets credit limits.
ČSOB and KBC Asset Management — ČSOB and KBC Asset Management debuted Enhanced Intelligence, Slovakia's first AI-managed investment fund. Nearly 200 mathematical models process over 1,200 economic and market indicators spanning 25 years of history, an algorithmic asset management approach where a human expert signs off on each decision.
Credit Agricole Bank Polska — Mizzox, Credit Agricole Bank Polska's new AI digital assistant for business banking, now sits inside the CA24 Mobile app for corporate customers. Natural-language search powers the tool, letting users pull up documents, draft reports and track projects through simple queries. Subscribers reach it via basic or advanced tiers under the bank's Klub Korzyści rewards scheme, one variant of an AI-powered virtual assistant for business clients.
KB Insurance — South Korea's KB Insurance runs an AI agent car accident fault assessment tool, calculating fault ratios for reported crashes via automated liability determination instead of manual guideline checks. Called the Automobile Accident Fault Ratio AI Agent, the generative AI system reads claim details and assigns splits alone, speeding claims and payouts.
AtlantaSanad Assurance — AtlantaSanad Assurance rolls out Assia, Morocco's first AI insurance assistant, staffing the insurer's site 24/7 to resolve queries instantly. Claimants get guidance on car accidents and the friendly accident report, a list of needed paperwork, and notes on contract guarantees and coverage caps. Assia covers subscriptions, claims handling and buybacks too, marking Morocco's debut round-the-clock virtual insurance assistant.
HSBC — HSBC Private Bank staff in Singapore and Hong Kong use Wealth Intelligence, a generative AI private banking tool on OpenAI's LLM. It sifts 10,000-plus outside feeds, sharpening how fast relationship managers spot market shifts and craft client-specific plans. HSBC will extend this AI wealth management platform elsewhere.
Westpac — Westpac has deployed Microsoft Copilot Studio organisation-wide for AI agent development within a governed setup, letting employees build their own generative AI agents. Dashboard queries via natural language are already possible through assistants embedded in SharePoint.
MoMo — MoMo has rolled out iProov Dynamic Liveness to tens of millions of users in Vietnam, applying AI fraud detection in fintech against deepfakes and digital-injection attacks. This makes MoMo iProov's first national partner and meets State Bank of Vietnam Decision 2345's call for stronger biometric identity verification for payments.
MAE Assurances — A sector first: MAE, Tunisia's Mutual Insurance of Education (Mutuelle Assurance de l'Education), has switched on an AI insurance agent reachable 24/7 through its app and site. Members get instant replies on coverage and step-by-step help with paperwork from this AI-powered conversational agent for insurance. The tool supplements human advisors rather than replacing them.
ACBS (ACB Securities) — ACB Securities (ACBS) embedded an AI stock analysis assistant, SMARTY, in its ACBS SMART app for Vietnamese clients. The AI investment assistant runs on Amazon Bedrock and Anthropic's Claude 3.7 as a multi-agent setup, mining reports to pitch trades, field questions and steer tactics free, geared toward retail traders over fund managers.
Direct Pojišťovna — 30% of windscreen claims at Direct Pojišťovna now close with zero human input via AI claims automation insurance. Artificial intelligence in claims processing drafts the file for the other 70%, cutting staff work to near nothing. Speed: 2 minutes per claim, down from 15, non-stop.
TBC Bank — TBC Bank debuts Georgia's first generative AI banking chatbot, built in-house. The assistant parses customer questions and gives thorough financial guidance, with skills set to grow through ongoing learning. Live now on the lender's website, the AI virtual assistant for banking should later reach every TBC digital channel.
Kuwait Finance House — Kuwait Finance House (KFH) launched staff tools for AI in Islamic banking. ShariaBot, a virtual Islamic finance expert, answers Sharia-compliant banking queries citing KFH's Sharia Supervisory Board fatwas; ComplianceBot covers Central Bank of Kuwait regulation and licensing. AI Copilot organises daily tasks; Speech Analytics transcribes calls; Automated Quality Management scores customer calls against predefined standards. Islamic banking AI adds a smart data analytics initiative.
SBI Insurance (SBI Ikikai Shoutanki) — Callers to Miyazaki Business Centre are now greeted by an AI voice bot insurance call centre system built for SBI Group's insurer SBI Ikikai Shoutanki Hoken, resolving over 70% of inquiries unaided. Mobirus's agent-type voice engine powers this AI customer service in insurance, trimming handoffs to staff by 63%.
MUFG Trust Bank — Mitsubishi UFJ Trust Bank now runs ASSIST, a Kasanae-built generative AI assistant retail banking tool, across roughly 2,000 staff in Japan who consult it daily on manuals, rules and procedures. The system shortens time otherwise lost searching for answers and handling queries. This AI staff support tool for banking operations is projected to yield savings near six thousand five hundred working hours a year.
National Bank of Ethiopia — AI transaction monitoring in banking is now catching irregular activity across Ethiopia, credited to the National Bank of Ethiopia under Governor Mamo Mihertu. This AI-powered fraud detection system, spanning lenders and fintech operators, has already frozen 138 accounts linked to suspicious dealings. The rollout, officials say, feeds into a broader campaign against illicit cross-border transfers and terrorism financing.
Samsung Fire & Marine Insurance — Samsung Fire & Marine Insurance in South Korea now applies AI claims review medical insurance processing, automating cancer diagnosis checks and surgery benefit payouts. Generative AI with optical character recognition scans records up to 48 pages, extracting data from diagnosis paperwork and lab results without manual work. The insurer reports faster, more accurate claims processing.
Community Bank Bangladesh PLC — MSME lending in Bangladesh now runs through an AI credit scoring for MSMEs tool built by Community Bank Bangladesh PLC with partners PriyoShop and Insights Genie, the first alternative scoring model of its kind in the country. Rather than relying on collateral, it weighs demographic data, purchase patterns, psychometric signals and behavioural analytics to judge creditworthiness, opening short-term loans to small retailers long excluded from formal banking via a fast eKYC process. This small business credit assessment marks a shift from conventional collateral-based lending.
African Bank — African Bank's payment channels now run on INETCO BullzAI, an AI transaction firewall supplied by NCR Atleos, catching fraud the instant it occurs. Models retrain continuously after each transaction, letting the fraud detection system flag emerging schemes early.
IAG — Insurance Australia Group runs AI property claims assessment on every claim, flagging urgent make-safe repairs after a hazard via generative AI. Automated review replaced contractor site visits, saving hundreds of thousands of dollars and beating manual accuracy. IAG applies this damage-assessment approach fleet-wide.
Banco de Occidente — Banco de Occidente, with AWS and Aval Digital Lab, built ARI (Intelligent Responses Assistant), an AI customer service banking tool checking PQRS replies against nine criteria. Sampling rose from 5% to 36% across 4,072 responses, review time fell to 1.3 minutes (17.3x faster), 95% compliance under this AI-assisted support process.
Cornerstone Insurance — Nigeria's Cornerstone Insurance Plc deployed Cici, an AI chatbot in insurance, spanning several digital channels. Policyholders now secure pre-priced cover and resolve claims via the generative AI-powered chatbot, capable of managing numerous enquiries together through adaptive, situation-aware exchanges instead of scripted replies.
Iute Macedonia — Iute Macedonia, within Estonia's Iute Group, now runs North Macedonia's first fully automated loan decision engine. The AI credit approval system blends official institutional records with machine learning risk segmentation, shrinking turnaround on consumer and earmarked loans from 15 minutes to 3, an early instance of automated lending needing barely any staff input.
DNB — Norway's largest bank is extending Microsoft Copilot to 5,000 staff, adding 3,000 more employees after a trial among 2,000 found 95% using the assistant weekly. DNB frames this AI deployment in banking as building internal skills through wider AI tool rollout in banking rather than chasing quick, quantifiable gains.
Bank of the Philippine Islands — Bank of the Philippine Islands (BPI) gave its 24/7 AI banking chatbot, BEA Chat, a real-time service request tracker, product applications and personalised replies within three seconds. Via BPI's website and Facebook page, the conversational assistant fields banking questions and self-service requests, routing complex cases to live agents.
Cámara de Aseguradores y Reaseguradores de la República Dominicana (CADOAR) — CADOAR (Cámara de Aseguradores y Reaseguradores de la República Dominicana) has launched Cadia, an AI assistant for insurance sector use, built on CrecerLab's Lisa platform. Cadia answers real-time information requests from insurers, brokers, regulators and the public, modernising access to insurance-sector data in the Dominican Republic. The rollout gives the industry a shared conversational AI channel for enquiries once handled manually.
Kifiya Financial Technology — Kifiya Financial Technology's AI credit scoring has unlocked over 35 billion Ethiopian birr in collateral-free financing for 467,700 micro, small and medium-sized enterprises, backed by the Mastercard Foundation. Alternative data and predictive analytics replace conventional credit history, reaching 355,027 women entrepreneurs and sustaining a million jobs. This algorithmic loan assessment aids collateral-free capital access.
BBVA México — Touch-tone IVR menus are gone at BBVA México, swapped out for Blue, a generative AI assistant fielding customer calls in natural language across 63 million annual calls — an AI customer service in banking overhaul. The bank claims this generative AI virtual assistant for call centres picks up caller intent correctly over 90% of the time immediately, trimming resolution from four minutes down to under sixty seconds, though the numbers remain BBVA's own and independently unaudited.
Tokio Marine Insurance UAE — Tokio Marine Insurance UAE now registers claims 90% faster after deploying AI document automation insurance with Kodak Alaris. Microsoft Document AI drives the Info Input platform behind this, tripling daily capacity while manual exceptions fell 60%. Motor policies make up most of the roughly 600 claims handled daily in this AI-driven claims workflow.
SMBC (Sumitomo Mitsui Banking Corporation) — An AI CEO chatbot answering as Sumitomo Mitsui Financial Group's Tatsuru Nakajima now handles staff questions, drawing on his past remarks via OpenAI's GPT-4o and RAG to mimic his reasoning as an AI-powered executive assistant. A Microsoft-based avatar version is still being built. SMBC calls it a route to normalising generative AI among staff.
ZhongBang Bank — ZhongBang Bank, an internet transaction lender in Wuhan, now uses an AI smart marketing assistant powered by large language models to serve WeCom customer engagement. Chats are parsed automatically, matched to intent, and answered with recommended knowledge and rewritten replies, cutting labour costs by 25% while raising response speed 20% and operational efficiency nearly 30%, the bank said, having shown it at the 2025 World Artificial Intelligence Conference in July.
FirstBank of Nigeria — FirstBank of Nigeria now offers Fibani, an always-on AI chatbot in corporate banking built into its digital platforms. The virtual banking assistant opens accounts, checks balances, moves funds, fields loan queries, logs ATM dispense errors, and lets users freeze accounts or blacklist USSD for security.
United Bank for Africa (UBA) — LEO, UBA's AI chatbot in banking, now settles cross-border transfers through the Pan-African Payment and Settlement System (PAPSS), a continental first. Funds move instantly in local currency between African nations via the chatbot for banking transactions, with no separate system required.
Wells Fargo — 215,000 Wells Fargo staff are gaining AI agents in banking through a widened Google Cloud tie-up centred on the Agentspace platform, part of a broader push toward agentic AI for banking. Contract work, foreign exchange questions, customer support and internal research now run through these tools. NotebookLM and Deep Research launched first with roughly 2,000 people, ahead of a phased wider rollout.
Generali Česká pojišťovna — AI property valuation underpins Generali Česká pojišťovna's new tool, Můj majetek 2.0. Map data and street-view imagery let the system estimate built-up area and roof style, checked against land registry records to set a fitting insured sum. Planned upgrades to this AI in building insurance underwriting will extend recognition to chimneys, solar panels, roof windows, and building storeys.
Ethniki Asfalistiki (National Insurance) — Ethniki Asfalistiki now settles claims in 2-3 minutes, running AI claims processing in insurance with Archeio8iki on the AI Flows platform. Vision models, OCR and language models scan medical files, sort them and pull the fields, an automated insurance document workflow. High-confidence cases skip straight to payout.
Commercial International Bank Egypt (CIB) — AI in lending now delivers automated credit decisions for new customers at Commercial International Bank Egypt (CIB), one of three live deployments there. InsightGini, a bespoke build on Dataiku, trims report effort 70% via a fine-tuned LLM; a separate tool ranks contact centre callers by sale likelihood, turning the unit profitable and cutting effort on targeted calls 80%.
Bank SinoPac (Yongfeng Bank) — Yongfeng Bank's AI agent in banking, "Yongfeng iWish," spans all 125 branches in Taiwan, taking spoken, typed or photographed deposit, withdrawal, transfer and remittance requests. By December 2025, this conversational banking tool tallied 80,000 transactions (8万), hit 95% satisfaction, cut counter time 30% and voice waits 70%.
Policybazaar for Business — ClaimSetu, built by Policybazaar for Business, chases a 50% quicker group health reimbursement cycle through AI in health insurance claims. NLP and OCR sift paperwork, catch mismatches and flag fraud while scoring submissions instantly, letting corporate insurers cut delays with AI-powered claims processing.
Univé — Univé now runs AI car damage claims through SchadeCheck, letting Dutch motor policyholders check in under 10 minutes whether a small claim is worth filing. A photo taken in the Univé app feeds the system, which calculates likely repair bills and flags the hit to no-claims discount before anyone commits to claiming. No other insurer in the Netherlands has paired this kind of AI damage assessment with photo recognition, and Univé is starting with the 10,000 to 15,000 single-party cases it processes yearly.
Royal Bank of Canada — ATOM (Asynchronous Temporal Model), Royal Bank of Canada's proprietary foundation model, now powers credit adjudication and personalised Avion Rewards suggestions for 1.3 million monthly users, marking a live case of AI in corporate banking. The system draws on billions of client transactions through RBC's Lumina data platform. This machine learning in financial services rollout anchors RBC's targeted $700 million to $1 billion in AI-driven value by 2027.
Jyoti Bikas Bank — To mark 18 years in business, Jyoti Bikas Bank rolled out Jyoti, an AI chatbot for banking customer service, reachable any hour via a simple, multilingual interface. It answers account questions instantly, widening the lender's digital service footprint in Nepal.
die Bayerische — Dental policies are the proving ground for German AI insurance application support: die Bayerische's assistant LEA steers buyers through finishing forms on screen. Startup muffintech co-built the system, next expanding into household cover. Examiners scored LEA 98% on an IHK-style exam, calling this digital insurance application tool muffintech's first venture into active selling over mere support.
Axis Trustee Services — AVA (Axis Trustee Virtual Assistant), built by Axis Bank subsidiary Axis Trustee Services, marks India's first AI chatbot for trustee services. The assistant resolves product and service questions instantly, tracks pending requests, and clears up transaction issues touching paperwork, regulatory rules and compliance checks — an early example of an AI-driven fiduciary support tool in this sector.
Bank of Baku — Round-the-clock digital support for Bank of Baku customers now comes from BOBAI, a newly launched AI virtual assistant customer service tool hosted on bankofbaku.com and Bolkart.az. Available 24/7, it handles queries on products and services while routing trickier issues to human staff. The rollout marks another step in the lender's push toward full digital transformation in Azerbaijan, extending its AI chatbot for banking support beyond typical business hours.
United Bank for Africa (UBA) — United Bank for Africa's AI chatbot LEO now handles instant cross-border payments for customers in Cameroon, an African first for AI-powered virtual assistant money transfers between countries. It runs on the Pan-African Payment and Settlement System (PAPSS), settling in local currency within seconds and sparing users the multi-day waits and higher costs of older channels.
ING Türkiye — ING Türkiye posts an 85% success rate in business processes through a bank-wide AI in banking operations rollout, active since early 2025. Generative AI in banking now reviews documents, wins new customers and handles back-office duties, letting employees focus on richer client relationships instead.
Migros Bank — Several AI assistants now operate live at Migros Bank across customer and staff workflows, Thomas Greuter said, marking a shift toward AI customer service in banking that answers general enquiries continuously. Contact staff instead get drafted email replies from a separate tool, reviewing each before sending. A daily-refreshed, curated knowledge base keeps accuracy high and limits hallucinated output — the bank's AI assistant for banking enquiries in practice.
Bradesco — Bradesco has launched Pix by voice on WhatsApp, driven by generative AI assistant BIA, enabling customers to move up to R$300 through spoken instructions or typed chat. A trial group of 1,000 clients in Brazil received early access, before a wider voice banking rollout scheduled for year-end.
Zurich Malaysia — Zurich Malaysia's Z Buddy answers agents' questions via WhatsApp on travel and motor cover, claims steps and product details. This AI insurance assistant speeds up routine tasks and lightens the load on Zurich's distribution network. As a virtual assistant for insurance agents, it frees staff for selling and servicing policies.
DSK Bank — DSK Bank rolled out Bulgaria's first fully AI-based assistant, an AI banking assistant driven by OpenAI's generative AI and trained on contextual data about its own offerings. Available via the bank's website and dskdirect.bg, it answers queries on products, charges and fees, an early AI-powered banking assistant example locally. Plans call for it to become a personal financial advisor giving recommendations and completing tasks with little staff input.
Kooperativa pojišťovna — Kooperativa Insurance and MAMA AI built the Czech Republic's first fully automated agentic AI claims processing life insurance system, needing zero human input between the initial report and the eventual payout call. Easy cases finish in roughly 12 seconds typically, whereas trickier ones get bumped over to a human adjuster instead. A homegrown composite model, custom-trained, powers this shift toward automated life insurance claims handling.
ERGO Versicherung AG in Österreich — ERGO Versicherung in Österreich now bases motor costs on WTW's Radar Live, an AI pricing platform for motor insurance and its main engine. Live data feeds algorithms weighing flexible excess and coverage variants per driver — pushing the insurer toward AI-driven premium calculation.
Allied Irish Banks (AIB) — Roughly 10,000 AIB staff now have Microsoft 365 Copilot built into Outlook, Word, Excel, Teams and PowerPoint, an early step in AI in banking. A dedicated AI Centre of Excellence develops bespoke tools using Copilot Studio, and coders will soon receive GitHub Copilot to speed development.
Faisal Islamic Bank of Egypt (FIBE) — Faisal Islamic Bank of Egypt (FIBE) switched on an AI-powered digital banking platform, Intellect Design Arena's eMACH.ai Digital Engagement Platform, bundling chatbots, self-service and three-minute onboarding. Retail customers get it first; SME and corporate clients follow as coverage widens, pointing to broader automated finance in Egypt.
Wiener osiguranje — Wienka, Wiener osiguranje's new AI insurance chatbot, gives Croatian customers 24/7 access to policy details, claims status, debts and agent contacts. Billed as the market's first such virtual assistant for insurance, it answers routine questions instantly, cutting the need for phone calls or branch visits.
Erste Bank Hungary — Erste Bank Hungary now answers complex retail banking questions in seconds via an in-house, OpenAI-based AI assistant in banking spanning branches and call centres. Drawing on documents covering policies, announcements and terms, this AI banking assistant frees staff from manual lookups toward customer relationships.
Banco de Crédito del Perú (BCP) — Banco de Crédito del Perú wired generative AI into Clara, its WhatsApp helper, a Peru banking chatbot leap dated July 2025. Clara reads urgency or frustration, catches a stolen-card alert needing swift handling, and untangles messy, open-ended loan or account questions with smoother, humanlike answers. BCP already topped Peru's lenders in running AI via phone banking, a generative AI banking assistant first this launch extends.
Commercial Bank of Ceylon — Commercial Bank of Ceylon unveils Sri Lanka's first AI-powered SME credit underwriting system, engineered with a global consultancy. Predictive analytics now drive faster, tailored loan decisions for small firms, marking a fresh use of AI in commercial banking.
Land and Houses Bank (LH Bank) — LH Bank built LISA, an AI chatbot for banking staff, using Sertis technology hosted on AWS. Workers now get faster answers on internal procedures, plus help with task management, cutting time spent hunting for information across departments through this workplace AI assistant.
CaixaBank — Card holders using CaixaBank's mobile app can query a generative AI agent for product exploration, comparing options and recommending a match. Google Cloud supplies the technology; 200,000 customers gained access in the initial card-products phase, with this AI-based financial product recommendation tool anchoring the bank's 2025-2027 Strategic Plan for generative AI.
TD Bank Group — TD Bank Group has piloted the TD Securities AI Virtual Assistant, a proprietary generative AI chatbot for institutional sales, trading and research staff answering clients faster. It uses retrieval-augmented generation and text-to-SQL to scan internal research documents, an AI virtual assistant in banking for internal workflows, not direct customer contact.
FinecoBank — FinecoBank's Portafoglio AI, built into the X-Net platform, turns a five-day process into under a minute. It delivers AI investment strategies for financial advisors: automatic projected returns, backtests and commentary, packaged as an AI-driven portfolio construction report, no manual work needed.
HNB General Insurance — HNBGI NEXA, an AI chatbot for insurance customer support, now runs on HNB General Insurance's site. Indian InsurTech firm AccelTree Software built it for round-the-clock routine queries, forwarding harder cases to the hotline.
Magyar Nemzeti Bank (MNB) and GIRO Zrt. — Hungary now runs real-time AI fraud detection in banking via the Central Fraud Detection System (CFDS), built by Magyar Nemzeti Bank and GIRO Zrt. to flag risky instant transfers within seconds. Lenders feed it payment data and fold its scores into their defences, anchoring wider continuous payment fraud monitoring.
Kuwait Finance House — Kuwait Finance House (KFH) has installed Fahad, the country's first AI-powered virtual employee, in a first phase at its The Avenues mall branch (Phase Four). Available around the clock, this digital bank employee explains banking services, fields customer questions and helps customers with ATM use.
Maib — Maib built ami, an AI virtual assistant for business clients, replying non-stop rather than routing queries to staff. Corporate users in Moldova reach it via website chat, Viber, Telegram and Facebook, bypassing the 1314 contact centre or a branch. This AI-powered chatbot for corporate customers was made in-house, not bought from a vendor.
I&M Bank — Kenya's I&M Bank now runs ThetaRay's cognitive AI anti-money laundering platform group-wide, covering Tanzania, Rwanda, Uganda and Mauritius too. One AI-driven engine now handles account opening, sanctions checks, monitoring, alerts and filings to regulators, swapping out older, fragmented compliance software.
BSP Financial Group — BSP Financial Group, Papua New Guinea's largest bank, has adopted SmartStream Technologies' Affinity module, an AI-driven tool for exception management in banking using observational learning to support automated exception management and data-based decisions. Nuni Kulu said the change cut manual work and lifted accuracy for over two million customers across PNG and the Pacific.
Eurobank Cyprus — Cyprus now hosts what Eurobank terms Europe's first enterprise-grade agentic AI voice platform, developed alongside Fairfax Digital Services and Voicing.AI. Staff hold lifelike spoken exchanges with these AI voice agents, which resolve IT questions and run operational chores unaided. Eurobank next plans wider adoption spanning HR, compliance and customer-facing roles.
First Abu Dhabi Bank — FAB now runs AI debt management on Intellect's eMACH.ai Lending platform, tying over ten systems together for live credit-book oversight and automated collections. First Abu Dhabi Bank sees it as one step toward AI-driven loan collection, aiming for full automation within five years.
Zürcher Kantonalbank (ZKB) — Zürcher Kantonalbank now lets ZKB ChatGPT, its in-house AI tool, handle harder tasks beyond simple research and drafting: building credit profiles for major corporate borrowers. Structured Finance staff feed the system data lifted from annual reports so it can produce a first-pass profile, an early instance of AI credit profiling, before specialists check the result. This AI in credit processes runs on Microsoft Azure and reached broader departmental rollout following a two-month pilot.
Bank of Georgia — Georgia's Bank of Georgia has rolled out Smart Loan, billed as the region's first AI-powered SME loan for small and medium enterprises. Approval takes under two minutes, with the system generating tailored financing offers automatically rather than through manual underwriting. The bank presents this as a shift in how business borrowers in the country obtain rapid access to credit through AI in SME lending.
Banco do Brasil — Banco do Brasil has built an AI personal finance manager into its Minhas Finanças Multibanco app, giving retail customers tailored credit line and investment tips from their financial profile. It claims to be Brazil's first bank offering AI-driven investment suggestions personalised this way.
Gránit Bank — Gránit Bank's new AI agent Gránit Guru now handles AI mortgage loan processing alongside staff, reviewing documents and applications. Automation should climb to 80-90 per cent, well past the 30-40 per cent reached through earlier digital upgrades. This deepens Gránit Bank's shift toward AI-assisted lending in Hungary.
Banco del Austro — Instant credit approvals, once a two-hour wait, are now standard at Banco del Austro after it rolled out a new IBM data platform for storage, integration, governance and lineage. Data engineer output has climbed, readying the Ecuadorian lender for broader artificial intelligence in banking through this AI-driven data platform.
Westpac — An AI scam detection tool now being trialled by Westpac transcribes live calls between customers and its fraud support staff, surfacing risk indicators mid-conversation. Prompts steer operators toward sharper, faster responses. Westpac frames this real-time fraud call analysis as a route to quicker fraud disruption.
Oromia Bank — Oromia Bank's Miliki app grants collateral-free loans up to 300,000 birr via AI credit scoring, weighing spending patterns, financial history and non-traditional data. Built with fintech partner Quantum Technology, it extended credit to over 7,000 borrowers during its trial run, evidence of strong latent demand for AI-driven lending among women, youth and informal workers excluded from conventional finance.
PrivatBank — PrivatBank launched AI counterparty checks for Ukrainian business clients via the web and Privat24 for Business app. The tool scans public records — ownership, network ties, financial standing — flagging legal and reputational risks before deals close: AI business reputation screening for corporate due diligence.
Markel Canada — Markel's Canadian underwriters gain agentic AI for insurance underwriting under an expanded mea Platform deal, following London Wholesale's earlier rollout. Unstructured submission data becomes structured through the platform's agentic GenAI, sharpening decisions — a use some label agentic GenAI for underwriting.
GFH Financial Group — GFH Financial Group now offers the GCC's first fully voice-operated AI-powered investment assistant, letting customers speak directly to manage holdings instead of tapping through menus. Running on OpenAI's technology, the tool handles live portfolio tracking and real-time investment inquiries. This puts GFH ahead of rivals still limited to text-based conversational banking tools.
Nasfund — Nasfund, Papua New Guinea's biggest private superannuation fund, runs AI invoice processing via DeepDelve Invoice AI, hitting 97 per cent data extraction accuracy on trialled supplier invoices. Innovation PNG Awards gave Nasfund its Large Business prize as the tool lifted productivity 40 per cent and freed three hours daily per worker. Rajeev Sharma says large language models come next, automating payments entirely, targeting a 70 per cent cut in processing time.
MBH Bank — Hungarian SMEs can search state grants via MBH Bank's Forrás AI assistant for SME grants on the BUPA platform, built with Peak Fintech. It turns dense legal wording from GINOP and DIMOP schemes into plain summaries, checks eligibility, and tracks funding-rule changes as this AI-assisted grant navigation covers micro-enterprises too.
DenizBank — DenizBank, Turkey's fifth-largest private bank, now launches new microservices models in 10 minutes, down from several days, using Red Hat OpenShift AI for AI model development in banking. About 120 data scientists there cover risk, marketing and customer relations, running machine learning lifecycles on-premise and in public cloud.
E.SUN Bank (Yushan Bank) — E.SUN Bank runs "GENIE" internally and gives retail customers "Jin Mishi", an AI in retail banking chatbot for queries. Its coders use GitHub Copilot, cutting build times 30% per the bank's own figure. All three sit inside a wider NVIDIA partnership spanning four finance AI use cases, including AI-assisted coding.
Amana Cooperative Insurance — Riyadh-based Amana Cooperative Insurance has deployed agentic AI for motor claims, running the whole journey through a swarm of master-and-worker AI agents teamed with RPA bots, leaving core-system code entirely unchanged. Built by eData, the setup carries a case from first notice of loss all the way to settlement and fraud screening, while customers simply reach out over WhatsApp or email instead of calling in. The insurer points to quicker paperwork turnaround, reduced leakage via automated cost checks, and happier customers thanks to round-the-clock chat availability — an early example of AI-driven claims automation in the region.
KB Financial Group — KB Financial Group has switched on KB GenAI Portal, an agentic AI banking platform giving staff across its eight subsidiaries the tools to build and run their own AI agents for financial services—a first for South Korea's financial industry. No-code and low-code options let non-programmers build agents too, alongside a pro-level track. Roughly 90 agents across 17 business lines, covering wealth management, retail and corporate banking, are due within three years, with advisory, client relations and support agents already underway.
Mutual Trust Bank PLC — Mutual Trust Bank PLC (MTB) became the first lender in Bangladesh to roll out an AI voice contact centre agent, letting account holders speak commands for balance inquiries, card activation and PIN generation rather than punching keypad options or sitting on hold. Available 24/7, the tool sharpens its own replies over time via machine learning, functioning as an AI virtual assistant for banking that keeps improving with use.
Boubyan Bank — Boubyan Bank built Kuwait's first home-grown AI model reading the Kuwaiti Arabic dialect at high accuracy, powering its AI call centre analytics. Speech analytics turns recorded voice calls into searchable text, mined for customer sentiment and actionable insights. Encrypted local servers safeguard customer data and process survey and social media feedback.
Sampath Bank — Five taps on the Sampath Vishwa Mobile App now trigger a full AI-powered personal loan decision, under Sampath Bank's new Sampath Select scheme, marketed as Sri Lanka's first fully automated lending product of its kind. Real-time analysis of applicant data replaces the old paper-heavy approval chain, producing an instant, paperless outcome. The pitch centres on AI-driven loan approval, with credit checks and sign-off compressed into one digital flow.
ColendiBank — Turkey gained its first AI-based digital bank as Colendi's ColendiBank opened to customers, having cleared licensing with regulator BDDK in October 2024. Its AI-driven backbone serves individuals and SMEs via embedded finance, tailoring products per user for AI in retail banking.
Cathay Insurance (Guotai Property Insurance) — Four new patents mark Cathay Insurance's (Guotai Property Insurance) push into AI in insurance claims processing across Taiwan: claims reports now write up twice as fast, loss-prevention reports take over 50% less staff time, and new tools handle underwriting risk monitoring and reinsurance exposure forecasting. A human-in-the-loop setup governs every AI-assisted risk assessment function.
Bank of Ayudhya (Krungsri) — Cloud migration sped up over 50% at Krungsri (Bank of Ayudhya) after adopting generative AI agents on AWS. AI fraud detection banking runs via Amazon SageMaker's AI-powered mule account detection; a valuation tool trims property visits from three days to 15 minutes; and machine learning balances cash across 5,000 ATMs nationwide.
Flow Uganda — Flow Uganda now runs AI sales automation in banking via Trembi Sales AI, scrapping 80 agents once posted across 30 districts. The Ugandan lender identifies and wins bigger loan prospects, restaurants and hotels included, at lower cost per signup. This AI-driven client acquisition for lenders swaps legwork for automated, high-value targeting.
Commerzbank — Ava, Commerzbank's AI banking avatar, handles routine queries and account tasks for German customers via the bank's mobile app, built on Microsoft Azure AI technologies. It can also act directly, ordering or blocking a credit card. Commerzbank calls this rollout one of Germany's earliest generative AI customer-facing deployments paired with avatar technology.
Mitsubishi UFJ Bank (MUFG) — Mitsubishi UFJ Bank now runs FRONTEO's "KIBIT Eye" for AI fraud detection in banking, spotting cartel links and early misconduct hidden across huge text archives. Automated screening cuts human bias, letting compliance staff work from thorough, uniform results. The move backs MUFG's wider AI compliance monitoring drive for customer safety and openness.
Bakai Bank — Kyrgyzstan's Bakai Bank now issues cards in one minute via machines using AI face recognition in banking, powered by BakaiGPT. Biometric facial identification for card issuance replaces document checks. Bakai Business Forum hosted the debut, the country's first such rollout.
FirstBank of Nigeria — Humanoid machines welcome visitors at FirstBank's fifth Digital Xperience Centre in Lekki, Lagos, a sign of AI in retail banking taking physical shape. They respond to queries instantly, alongside self-service kiosks, contactless ATMs, video banking and a machine issuing ATM cards within three minutes, watched over by fraud detection powered by artificial intelligence in banking.
Latvijas Transportlīdzekļu apdrošinātāju birojs (LTAB) — Latvijas Transportlīdzekļu apdrošinātāju birojs (LTAB) deployed a virtual assistant bringing AI in motor insurance to Latvia's motorists, handling enquiries on compulsory motor third-party liability (OCTA) insurance. Linked to LTAB online services, the insurance chatbot explains 24/7 what follows crashes, policy validity and claims.
Singlife — Singlife's #BetterIdea Challenge hackathon produced this AI insurance assistant for financial advisors, live on Oracle Cloud Infrastructure. Staff retrieve product details instantly rather than by hand, powered by OCI Generative AI and AI Vector Search — a generative AI tool for insurance advisors speeding replies to client questions.
Industrial and Commercial Bank of China — A dedicated thousand-card computing cloud drives Industrial and Commercial Bank of China's rollout of DeepSeek's newest open-source model, folded privately into its "ICBC Intelligent Surge" (工银智涌) system — a first for major Chinese lenders. AI banking deployment spans 200-plus scenarios across 20 business lines. Ten-plus large language model banking tools alongside 2,000 legacy models back this expansion.
Tryg — Fault gets settled within minutes, not days, for nine of ten two-vehicle collisions once Tryg, Denmark's biggest insurer, runs its AI model for car damage claims. Built on 42,000 prior disputes and checked against 4,000 more, the automated fault-determination tool still routes thornier disagreements to staff.
AXA — AXA runs over 400 AI use cases in insurance in production across its value chain, backed by 2,000 specialists, over 900 in data science and engineering roles. AXA Secure GPT, a generative AI assistant for employees, reaches all 154,000 staff across 79 business applications. SmartInAXA, a RAG-based tool, is rolling out to French distributors.
Unitrust Insurance — UNIS, Unitrust Insurance's new AI insurance chatbot, offers 24/7 policy queries, quotes and claims handling. Policyholders use this self-service insurance bot to file and follow claims, buy motor cover, pay premiums and download documents unaided.
Generali Switzerland — Misrouted calls at Generali Switzerland fell 22% after Leonie, a Swiss German AI voicebot, took over sorting incoming lines. Enterprise Bot built this voicebot for customer call routing on Luware telephony, reaching 97-98% speech recognition accuracy and 3% higher availability. Generali next wants Leonie fluent across all Swiss national languages, plus self-service options for policy changes.
Iraqi Private Banks Association — Masrifuna, from the Iraqi Private Banks Association, now houses an AI financial chatbot to sharpen users' money habits and ease access to banking details. Ali Tariq named it an AI-powered virtual advisor, drawing on figures supplied by member banks and Germany's GIZ, with the Central Bank of Iraq and the European Union backing its rollout.
MBH Bank — MBH Bank launched a generative AI chatbot for loan queries about Munkáshitel, Hungary's state-subsidised loan scheme. The Munkáshitel AI Assistant, built with Peak Fintech on a retrieval-augmented generation (RAG) framework, guides borrowers through complex eligibility, loan-size and repayment rules. The loan chatbot lightens bank staff workload.
Ping An Life Insurance — Ping An Life has rolled out AI insurance claims processing alongside a tailored DeepSeek-R1 model spanning three business functions: AskBob, a sales-support tool for agents; a training system for staff; and an automated engine for handling complex claims cases. Engineers built the customisation in just seven days, from raw data processing through to production release. More than 90% of agents judged the revamped AskBob better than its predecessor on completeness, accuracy and practicality, while the claims engine, now embedded in the core system, lifted accuracy on complicated cases by 31% versus the earlier version — an early signal of how automated claims review can scale inside a large insurer.
PICC Property and Casualty — PICC Property and Casualty leads Chinese insurers in adopting AI in claims processing, running DeepSeek's full-size model on home-grown hardware. DeepSeek now sits in the firm's model base, auto-checking claims quality, pulling facts from paperwork and sharpening staff training. Underwriting, claims and investment uses are still developing, pointing toward wider claims automation powered by AI.
Bank of Baroda — Bank of Baroda now runs three GenAI banking tools as part of its generative AI in banking rollout: Aditi, a Virtual Relationship Manager, manages conversational interactions via digital avatars; GyanSahay.AI gives employees an internal knowledge base; ADI fields customer queries as a chatbot. All three deliver 24/7, multilingual help, part of a wider GenAI-powered banking tools push for sharper service and staff output.
BBVA — Blue, BBVA's AI banking assistant built on OpenAI's technology, reached 10,000 users first, then all customers within two months, handling account and card tasks through conversational AI in retail banking across roughly 120 functions. BBVA also gave 100 managers the tool, linked to 30,000 documents on products and processes.
Türkiye Sigorta — Türkiye Sigorta launched Bilge'ye Sor, an AI assistant for insurance customer service, on its website, mobile app and Pusula Plus platform. This generative AI helper uses natural language processing for claims tracking, provision checks and document requests, live 24/7 for customers, agents and field teams.
Banca Generali — Microsoft 365 Copilot now covers every employee at Banca Generali, part of a broader push into AI in private banking, and Personal Assistant tools already reach 60% of staff. A parallel effort is building generative AI for financial advisory, supported by an internal AI Ambassador scheme.
Cooeprouasd — Cooeprouasd, the Cooperativa de Servicios Múltiples de Empleados Profesores de la Universidad Autónoma de Santo Domingo, has launched Cooperativina, an AI virtual assistant for cooperative members in the Dominican Republic. It answers questions on loans, savings and other services and processes payments, available 24 hours a day via text or voice. The move points to a wider shift toward AI-powered banking assistants at member-owned financial institutions.
Sukoon Insurance — Sukoon Insurance now offers AI face scanning for life insurance, replacing the standard medical exam with a short videoconference clip. Built with One Call, the system works by reading vital signs from facial video, capturing heart rate, blood pressure and breathing rate to speed up approval in the United Arab Emirates.
PZU Group — Homeowners insured under PZU Dom now see payouts settled via AI insurance claims settlement within 24 hours, as PZU Group's Samoobsługa NEXT platform applies GenAI to confirm coverage and trigger disbursement. This marks Poland's debut of generative AI claims processing in a live workflow.
IAG — Roughly 500 CGU staff in Australia now rely on CASI, an AI claims assistant insurance tool from IAG, to locate policy clauses instantly rather than spending 20-30 minutes searching manually. IAG intends widening this generative AI in claims handling into additional divisions beyond CGU.
Faysal Bank — Faysal Bank's i-Faysal, an AI virtual assistant banking customers use round the clock, handles balances, transactions, bills, transfers and product advice. The lender calls it Pakistan's first such AI-powered customer service assistant, replacing duties once handled by staff.
77 Bank (Shichijushichi Bank) — 77 Bank (Shichijushichi Bank) runs AI mortgage screening from Mitsubishi Research Institute, clearing over 50 per cent of home loan applications instantly without staff review. The Japanese lender adopted automated home loan underwriting once the model reached stable accuracy, keeping 100% uptime since launch.
Sun Life Philippines — Sun Life Philippines now runs AI-powered underwriting in life insurance on ALLFINANZ SPARK, a SaaS platform built by Munich Re Automation Solutions. Predictive models instantly score fresh submissions and paper forms, shrinking turnaround while raising straight-through rates. Policyholders gain quicker, automated underwriting decisions plus smoother onboarding.
Tokio Marine Seguradora — Tokio Marine Seguradora runs AI car insurance claims from phone photos sent via WhatsApp. Cília Tecnologia's 3D model reads damage and returns a cost estimate in minutes, skipping inspections and speeding payouts. This AI-based repair estimation replaces manual appraisal.
Trygg-Hansa — Swedish insurer Trygg-Hansa built a homegrown AI crime risk alerts system, spotting odd spikes in claims data and texting SMS warnings to policyholders facing rising burglary danger. An internal AI unit paired anomaly detection with clustering to build it, already covering thousands of customers. Car theft and other claims follow next as Trygg-Hansa widens beyond burglary-pattern notifications.
Banco Ripley — Lukas launched as Latin America's first AI finfluencer for Banco Ripley, teaching Chilean youth saving, budgeting and credit via social clips. Pedro Juan y Diego and Wise Innovation Studios built the avatar with motion capture and 3D modelling, voiced in colloquial Chilean Spanish. This AI-powered financial influencer format targets younger audiences.
Banca Transilvania — Banca Transilvania now runs an AI chatbot in banking called Chat BT, a DRUID-made assistant embedded in its BT Pay app, becoming Romania's first lender to place such an AI virtual assistant for banking customers into live client contact. Machine learning drives its answers on app functions and offerings in Romanian, with handoff to staff when required.
Sejfa — A generative AI claims intake insurance tool called Smart Skada now handles Swedish home insurance customers for Sejfa, the digital service owned by Länsförsäkringsbolagen. Instead of scripted replies typical of reproduction-style chatbots, this system drafts original follow-up questions and answers while collecting damage details through conversation. Testing began in September 2024, and strong outcomes led to a full launch, with expansion planned across Länsförsäkringsbolagen throughout the year. The generative AI insurance chatbot marks a shift from rule-based automation toward dynamic, self-generated dialogue.
JPMorgan Chase built LLM Suite, a secure in-house generative-AI platform for idea generation, drafting and summarising, and onboarded 200,000 employees within eight months of its mid-2024 launch; American Banker named it 2025 Innovation of the Year. The bank reports over 230,000 staff now use it worldwide, saving several hours a week.
ERGO Versicherung AG in Österreich — All staff at ERGO Versicherung in Österreich now use ERGO GPT, an in-house generative AI insurance tool developed by ERGO Group, after a trial period and required training for early users. The AI insurance assistant helps employees organise information, draft written communications, plan meetings, brainstorm ideas and manage everyday administrative tasks.
China Construction Bank — China Construction Bank has built DeepSeek-R1 into its own systems, adapted since February 2025 using internal records. AI in banking spans half the workforce over 46 divisions and beyond 200 jobs: scoring risk, converting payments, writing summaries, checking code. Review time per client dropped from 30 minutes to five under this model effort.
Baiduri Bank — Baiduri Bank claims Brunei Darussalam's first generative AI banking customer service launch, rebuilding Emmi from a declarative chatbot into a model handling complex queries with tailored replies. Unveiled at TechXpo 2025, it is also being piloted on WhatsApp, extending conversational AI for customer support beyond the bank's app.
Česká spořitelna — Česká spořitelna leads Czech banks in wiring OpenAI technology into its customer voicebot, an early case of generative AI in banking now fixing meeting times and firing off outbound calls unprompted. 40% of its roughly 35,000 monthly calls close with zero human touch, and the bank's own count puts bookings past 2,000 a month via this AI voicebot for banking, plus less phone time for staff.
Banco Nacional de Costa Rica — Banco Nacional, Costa Rica's largest state-owned lender, adopted AI cybersecurity tools between June 2024 and June 2025, sharpening detection of digital intrusions and speeding up its reaction to them. That upgrade helped judges award the bank its fifth Bank of the Year title, calling the AI-driven threat detection one of the strongest factors behind the win.
BNP Paribas Corporate & Institutional Banking — BNP Paribas Corporate & Institutional Banking now gives 43,000 staff and contractors secured in-house GPT access via LLM@CIB, part of a '1+5' push on AI in corporate banking. Noa, its portal-based agent inside Centric and Neolink, handles 100,000 clients' basic queries non-stop, spreading generative AI for banking staff into customer service too. Separate specialist systems back sales, research, IT development, operations and risk.
Signal Iduna — Signal Iduna has switched on Google Gemini Enterprise company-wide, letting German staff build custom AI agents via an internal system called Co SI — one of Europe's earliest full rollouts of generative AI in insurance. A narrower health-division assistant using AI agents for insurance staff already cut call transfers from 27% to 3% and raised processing speed by 37%.
Oman Housing Bank — Round-the-clock answers on subsidised loans now come from an AI chatbot for housing loans, deployed by Oman Housing Bank on its Iskan platform. A civil ID number lets applicants check waiting-list status instantly, sparing branch trips and call-centre queues. The bot service handles paperwork questions, 24/7.
Transbank — Eugenia, Transbank's new AI assistant for internal operations, now handles staff questions on time off, brand rules and compliance without waiting for a human reply. The Chilean payments firm built the tool on AWS language models as proof that automation can take hold inside everyday corporate workflows. Sebastián Robles, who leads AI at Transbank, said the goal was giving staff real systems to actually use, not just pilots. That practical rollout of an AI workplace assistant is now held up as a model for other Chilean businesses.
Aseguradora General — Aseguradora General added voice-message support to Sofía, its AI insurance quoting assistant in Guatemala, joining the existing WhatsApp and text channels in version 5.0. Customers now speak natural-language requests to get auto and medical expense insurance quotes in under a minute. This virtual insurance assistant has handled over 60,000 auto quotes and 5,000-plus medical quotes since debuting in 2024 as the country's first such tool.
Banque Internationale à Luxembourg (BIL) — Banque Internationale à Luxembourg (BIL) built Berry, an AI banking assistant answering card, payment and product questions from retail customers non-stop. Staff gain time for tougher cases. BIL enters conversational AI in retail banking.
Ryt Bank — Over 50,000 Malaysians tap Ryt Bank, a YTL Group and Sea Limited venture billed as the world's first AI-powered bank, logging some 80,000 transactions monthly. Talking replaces tapping: assistant Ryt AI, built on homegrown model ILMU, moves funds via plain speech, not menus. Chat is the interface in this AI-driven banking design.
NCB Capital Markets Limited — An AI investment chatbot named Simone now handles portfolio guidance for clients of Jamaica's NCB Capital Markets Limited over WhatsApp. Behind the scenes sits an AI-powered Risk Profiler and Asset Allocator, posing a brief round of queries to size up someone's appetite for risk before proposing a blend of stocks, fixed income and other instruments. NCB unveiled the system at the Jamaica Stock Exchange's Regional Investments and Capital Markets Conference.
Storebrand Bank — Storebrand Bank's generative AI chatbot handles 40% of customer traffic, answering queries and completing tasks like ordering BankID. This AI customer service banking deployment cut average wait times to 22 seconds and phone enquiries by 23%, while satisfaction rose 43% where the GPT chatbot was used. It stands as one of Norway's more detailed published rollouts of a virtual assistant handling routine banking requests.
Nest Bank — Photographing a bill now suffices for AI invoice payment through Nest Bank's app: the assistant N!Asystent reads the document, populates the transfer fields, verifies them and completes payment. Microsoft Azure OpenAI's GPT-4o powers the feature, reflecting Poland's move toward automated bill scanning and settlement.
novobanco — 599 autonomous AI agents now operate within novobanco, powered by Microsoft Copilot across branch, contact-centre and credit analysis functions, an AI banking deployment across Portugal since 2025. Over 1,500 employees trained while 60 use cases launched, trimming repetitive task duration by 75 per cent. Mortgage document review once over an hour now concludes within 20 minutes, part of a wider shift toward AI-driven financial services automation.
Banco Montepio — Searching millions of records now takes seconds at Banco Montepio, through an AI platform for internal data access named Pio, built by Devoteam on Microsoft Azure. Bots handle queries across credit risk, mortgages and sales, replacing manual lookups. Rules under the EU AI Act are already built in, part of a broader move toward AI in banking and financial services.
Sovcombank — Sovcombank's AI in banking push put over 80% of core staff onto daily AI assistants, with IT hitting 95% adoption. The Russian lender trained 34,000+ employees via a five-tier skills curriculum and launched 30 autonomous agent projects, one of the sector's largest workplace AI adoption drives.
PREMIUM poisťovňa — Slovakia's first AI advisory tool insurance intermediaries can use came from PREMIUM poisťovňa, built on OpenAI technology. Agents get replies in seconds from policy files, legislation and guidance, instead of hours of manual search. Humans still judge each case; the assistant for insurance intermediaries only shortens retrieval.
ICA Banken — Over 70,000 monthly sessions now flow through ICA's network of AI help robots, with 92% of users never needing to escalate to a live agent. Swedish retailer ICA built this AI chatbot banking capability first for its banking arm, ICA Banken, on icabanken.se, following earlier launches at Apotek Hjärtat in November 2024 and ICA Sverige the month after. This AI customer service chatbot expansion reflects a coordinated group-wide strategy rather than a standalone project, though the usage figures remain ICA's own and are not independently verified.
Allianz UK — Allianz Commercial deploys BRIAN, an AI underwriting assistant fetching answers from internal guidance so employees avoid combing through dense paperwork manually. Beyond its trial run, the system logged over 13,000 queries, saving roughly 65,000 minutes once lost chasing information — signalling AI-assisted document retrieval expanding across UK insurance.
The AA — Breakdown and insurance queries at The AA now go to boost.ai AI customer service agents, live on web within 30 days and on WhatsApp a month later. Over 90% of replies land accurately, and more than 45% of conversations never reach a human handler, thanks to this hybrid generative AI deployment.
Activo Banco Universal — Activo Banco Universal now runs round-the-clock AI customer service chatbots, built on natural language processing, fielding client queries 24/7. Tricky cases route straight to human executives, cutting wait times. Announced at IA Tech Day, the bank's next step adds chatbots grasping conversational context.
TBC Uzbekistan — AI voice agents now field over 50% of TBC Uzbekistan's payment reminder calls, the latest step in the Central Asian digital lender's AI banking automation push across Central Asia. Since January, a bilingual LLM built for product sales has driven Uzbek and Russian speech recognition and synthesis, with capacity flexing automatically alongside call centre demand. Staff at TBC describe these AI voice agents in banking as up to ten times stronger than human sales reps in certain instances, a claim from the firm itself rather than an outside audit.
InfinBANK — InfinBANK's AI credit scoring banking model gained 10% accuracy from new behavioural signals, part of an in-house AI stack built on Gemma and Llama, tuned for Uzbek and Russian. A hologram greets branch visitors; a Telegram bot answers queries. Its own data centre hosts the ML-based system, sharpening AI-driven customer risk assessment.
MB Bank — MB Bank (Military Commercial Joint Stock Bank) has halted hundreds of billions of Vietnamese dong in suspected fraud after rolling out AI fraud detection in banking beyond earlier pilot work. Vice Chairman Vu Thanh Trung pointed to automated screening as the driver, while AI-driven banking fraud prevention now supports retail sales staff and broader internal efficiency.
Shinhan Securities Vietnam — Shinhan Securities Vietnam embedded S-Era, an AI stock assistant, into its 'Sàn Xịn Ha' trading app, arming users with live pricing, forecasts, condensed news, equity analysis and valuations through charts or text, plus Q&A follow-up. This puts AI-driven stock analysis inside a Vietnamese brokerage's daily toolkit.
Absa Bank — Abby, agentic AI in business banking from Absa Bank, built on Salesforce's Agentforce platform, gives South African firms instant loan and investment guidance. Africa's first such rollout targets client service. Absa projects the AI banking assistant will double contact centre output.
FWD Life Insurance (Thailand and Cambodia) — FWD Insurance's generative AI voicebot and chatbot cut policy cancellations by 60% and lifted retention by 41% in Thailand. Both apply natural language understanding to read customer intent and emotional tone, with the chatbot giving sharper answers and the voicebot scoring high on satisfaction. This AI chatbot customer service upgrade relies on the insurer's own figures, not independently audited.
VietinBank — Built on Microsoft Azure AI Foundry, VietinBank's Genie assistant fielded 350,000 staff queries within two months, serving Vietnam's banking workforce. Drawing on 2,000 processes and 500 financial indicators, this AI in banking operations tool slashed response delays by 95%, sparing hundreds of thousands of monthly work hours through generative banking AI.
Western Financial Group — Insurance brokerage Western Financial Group has rebuilt its service operations on Quandri's AI. The companies credit the change with a 15% lift in personal-lines EBITDA and $2.5 million extra margin, a 2.3% retention gain worth $1.5 million in marginal revenue, and 14 full-time roles moved from admin to client relationships, across 180-plus branches and 300,000 personal-lines policies. No independent evaluation exists, and the release gives no measurement period.
Scotiabank — Artificial intelligence saved Scotiabank roughly 24,000 days' worth of work in about four and a half months, chief executive Scott Thomson has said, a headline for AI in Canadian banking. Bank of Montreal's Darryl White puts AI underwriting decisions in insurance at as little as 10 seconds. Royal Bank of Canada's investor day last year set a $700 million to $1 billion enterprise-value target by 2027. These bank AI productivity claims land as analysts weigh job impact.
Glassdoor — Glassdoor found 98 per cent of claims-adjuster reviews mentioning AI negative, the harshest among occupations studied. Adjusters said leaders pushed faulty tools whose mistakes staff had to fix, hurting service to policyholders. Adjuster employment fell 21 per cent from May 2025 to May 2026 (Bureau of Labor Statistics), and entry-level postings are down 50 per cent since 2025. Insurance was the third most AI-critical industry, with 81 per cent of comments negative.
Czech National Bank — A central bank AI unit is taking shape within the Czech National Bank's information technology arm under Janis Aliapulios, pooling capacity the institution already holds into one coordinating team and adding no posts overall. Governor Aleš Michl put the headcount reduction at 5 percent, attributing the departures not to poor performance but to the same work becoming achievable more productively — the governor's own account, with no published external assessment of any link to the in-house AI capability. Grok and OpenAI were tried against inflation forecasts during 2025, MiCA licence files now pass through such tools, and servers carrying Nvidia graphics cards have been purchased.
Sberbank — Sberbank expects AI in banking to deliver a 550 billion ruble economic effect across all business lines in 2026, CEO German Gref told shareholders. He linked AI adoption in banking to stronger revenue from personalised offers, sharper pricing, lower credit risk and tighter cost control, adding that AI agents already produce credit proposals for 96% of corporate borrowers.
Jubilee Holdings — Jubilee Holdings' AI fraud detection insurance push stopped KES 1.28 billion (Sh1.28 billion) in fictitious claims for the year ended December 2025, nearly triple the prior year's total. AI-driven claims adjudication tools, including Curacel, traced KES 750 million (Sh750 million) of that sum and flagged 449 suspicious cases. It sits within Jubilee's Changamk@ programme, cutting over 96,000 man-hours of manual work a year.
The US Treasury's Do Not Pay prevented, detected or recovered a record US$11.7 billion in fraud and improper payments in FY2025, up 63% from US$7.2 billion in FY2024, across 614 million data-source queries. It complements the FY2024 machine-learning check-fraud figure, though FY2025 reporting does not separate the ML-only line.
Allianz Türkiye — Allianz Türkiye's Smart STP, applying AI in health insurance provisions, has handled 100 million transactions. Automated insurance approval now settles 92% of provision decisions autonomously, clearing cases within seconds so patients no longer wait at hospitals. Allianz put 3 million euros into AI projects across three years.
Lunar — 85% of incoming customer queries at Lunar now get resolved by its AI customer service tool without human intervention, the Nordic challenger bank's best figure since the AI Support feature debuted in summer 2024. Satisfaction stays high too: 65% of those automated cases got a 4 or 5 out of 5 rating from users. This Denmark-based fintech's overall Trustpilot score, built on over 24,000 reviews, sits at 4.6 — evidence that AI in customer service can scale without hurting scores.
Machine-learning AI helped the US Treasury recover US$1 billion in fiscal year 2024 by detecting Treasury check fraud faster. The department's broader technology and data push stopped or clawed back a record US$4 billion in fraud and improper payments, versus US$652.7 million the previous year.
Magyar Nemzeti Bank — Over 170 AI banking use cases now sit in Hungary's banking sector: operations, customer relationships, sales. Izer Norbert, who runs financial markets and digitalisation at the Magyar Nemzeti Bank, told the AI Summit 90 percent of domestic lenders use artificial intelligence already or mean to within twelve months. Eight big lenders each carry five or more active AI projects.
UBS — Starting in 2027, UBS will reject graduate and intern applicants who cannot demonstrate AI skills in investment banking. Candidates must prove they can use AI tools to lift efficiency and results, a standard now placed next to the usual academic bar. UBS becomes one of the earliest banks to make artificial intelligence competency in banking a hiring gate rather than something learned on the job afterward.
Similarweb — Similarweb data shows AI-referred traffic to Latin American banks rising sharply between January and July 2026, from +128% in Mexico to +294% in Peru and Argentina. Santander Argentina saw the region's biggest jump; AI assistants otherwise sent most visitors to BCP in Peru, Bancolombia in Colombia, BancoEstado in Chile and Banamex in Mexico.
KPMG / Association of Serbian Banks — KPMG and the Association of Serbian Banks scored AI maturity in Serbian banking risk at just 28% across 18 lenders surveyed. Most rely on scattered, ad hoc pilots rather than a coherent roadmap for artificial intelligence in risk functions, held back by weak data readiness. Assessed dimensions: strategy, deployment depth, staff skills, risk controls, data quality, infrastructure.
FinTech Belgium, Sailpeak and Vlerick Business School — The third AI Barometer from FinTech Belgium, Sailpeak and Vlerick Business School sees generative AI in financial services reaching Belgian core processes beyond coding and customer service. 67% deploy it in claims processing and 46% in KYC/ALM, 64% pilot it in underwriting, and institutions seeking positive AI returns rose from 47% in 2025 to 75% in 2026.
Shaqiu Think Tank — Shaqiu Think Tank's tracking report finds 53.5% of Chinese banks, more than half, now using large models in banking, with the sector shifting from tool-level uses to system-level rebuilding of processes. The report does not give its survey base.
Association of Kenya Insurers — AI in claims management is shrinking Kenya's insurance payout waits to just 30 days, the Association of Kenya Insurers says. A health platform now processes daily claims 500% quicker, approvals landing in under sixty seconds, while chatbots deliver instant answers to customer questions at multiple firms. Fraud screening and actuarial modelling lean on machine learning as well — automated claims processing that counts, since AKI reckons almost 30% of submitted claims are bogus.
Financial Sector Conduct Authority and Prudential Authority — A joint survey by South Africa's Financial Sector Conduct Authority and Prudential Authority put AI in financial services at 10.6% penetration, based on answers from about 2,100 firms, with banks well ahead of other institutions at 52% uptake. Ethical oversight lags far behind: barely 13% of respondents deploying GenAI reported any governance safeguards, underlining weak generative AI adoption in banking despite growing usage.
Bank of Japan — About 50% of financial institutions surveyed by the Bank of Japan already deploy generative AI in banking, its FY2025 poll of 153 firms found. Counting pilots pushes that share past 70%, and including firms merely weighing adoption drives it beyond 90%. Officials cited swift AI adoption in financial institutions nationwide, yet flagged weak usage monitoring and third-party risk oversight.
Zoho (survey by) — AI adoption Nigeria is accelerating: Zoho's poll of 386 respondents put the figure at 93%, financial services leading at 29%. Over 50% have moved past pilots into full operational use, part of a wider shift in AI in Nigerian business. Customer service automation tops applications at 49%, ahead of software development at 46% and marketing optimisation at 32%.
Postbank (Eurobank Bulgaria AD) — Over 1.2 million people in Bulgaria have used Postbank's EVA AI banking assistant, which has logged more than 8 million queries to date. Usage is climbing fast: 320,000 exchanges came in just the first seven months of 2025, putting the full-year figure on track near 560,000. Reply speed now sits under 0.1 seconds thanks to a freshly deployed model, and the banking chatbot's topic range has widened past 800 areas since debuting in 2018 as the nation's first of its kind.
PwC Ghana — Ghana's bank chief executives are embracing generative AI in corporate banking, with 68% confirming use of the technology across 2024, according to PwC's 2025 Ghana Banking Survey. Chatbots now handle personalised customer interactions at several lenders, while staff lean on tools like ChatGPT to speed up everyday internal tasks. That level of uptake signals a clear shift toward automation spanning both front-line service and behind-the-scenes operations, a trend in AI adoption across banking that the survey highlights.
United Bulgarian Bank (UBB) — Autonomy at UBB's AI assistant in banking, Kate, climbed to 71% in 2025 from 54% when she launched. Customer engagement surged too: 762,573 people held nearly 2.2 million conversations with Kate from January to June 2025 alone, marking a 4.5-fold rise in users and 6.5-fold rise in exchanges versus her first year. Kate now performs work equal to 25 employees and moved 50,000 vignettes in 2025, while UBB Asset Management's AI-managed mutual fund attracted EUR 23.3 million from 597 investors.
Central Bank of Kenya — Credit risk scoring tops the list of AI in Kenyan banking, Central Bank of Kenya reports, cited by 65% of lenders already deploying the technology. Guarding against fraud follows at 54%, handling customers at 43% — though half of those polled had adopted none of it. Digital lenders far exceed traditional banks in applying AI-driven credit risk assessment, 80% versus 45%.
Finansforbundet — Eight of nine leading Danish banks now run generative AI in financial services at full scale, per a Finansforbundet poll. Home-grown GPT tools drive this: Danske Bank (DanskeGPT), Sydbank, Saxo Bank and Arbejdernes Landsbank each built separate systems for staff. Client-facing AI adoption in banking lags, though Lunar's assistant resolves 65% of written enquiries unaided.
Financial Supervisory Commission (FSC) — AI adoption in Taiwan's financial sector reached one in three institutions — 126 of 383 surveyed by the Financial Supervisory Commission — a jump from 29% in 2024 to 33%. Penetration was highest among banks (87%), then life insurers (67%) and property insurers (45%), while generative AI specifically had reached 61 firms. Back-office administrative work ranked as the dominant application, claimed by 39% of adopters, pointing to artificial intelligence uptake among Taiwanese financial institutions still weighted toward internal efficiency gains rather than client-facing services.
FINMA (Swiss Financial Market Supervisory Authority) — FINMA found half of some 400 Swiss institutions practise AI adoption in banking, live or piloted. A quarter more plan to start within three years; 91 per cent of adopters use generative tools too. Larger firms outpace smaller peers in application count, signalling growth in AI adoption in financial services.
Febraban — Generative AI in banking has reached eight of every ten banks across Brazil, per Deloitte's Febraban Banking Technology Survey 2025, with 95% using AI to some degree. Efficiency rose 11.4% on average from adoption in financial services, and 38% of banks saw gains topping 20%.
National Bank of Ukraine — AI adoption in Ukraine's financial sector has reached 64% of institutions, per a National Bank of Ukraine (NBU) survey of 208 regulated entities including banks and insurers. Of these, 23% described their use of AI or machine-learning tools as active rather than exploratory. Regulators pointed to the figures as proof the sector has moved from experimentation towards genuine practical use, with machine learning in Ukrainian banking now a working reality.
The Bank of England and FCA's 2024 survey of 118 firms found 75% of UK financial services firms already using AI, up from 58% in 2022. Another 10% plan AI adoption within three years.
National Bank of the Kyrgyz Republic — Kyrgyzstan's central bank charted a five-year path for AI banking supervision: the National Bank's Board approved the Concept and Roadmap for the Development of SupTech Supervisory Technologies for 2026-2031. Oversight shifts toward machine learning and big data, covering automatic data checks, online licensing and anti-money laundering screening, under wider supervisory technology adoption.
Central Bank of Kenya — Central Bank of Kenya wants prior sign-off from any lender letting models settle loan decisions, extending Kenya AI regulation to the credit desk. Its own 2025 count put half of licensed institutions on AI, while one tier-one player pushes Sh1.5 billion daily through machine-learning mobile lending, squarely inside Kenyan AI oversight.
Hong Kong Monetary Authority — The Hong Kong Monetary Authority, the Securities and Futures Commission, the Insurance Authority and the Mandatory Provident Fund Schemes Authority, with Hong Kong Cyberport, have picked 36 use cases from nearly 100 proposals for the first GenA.I. Sandbox++ cohort. 30 financial institutions and 27 technology partners take part, including Ant Bank (Hong Kong), WeChat Pay Hong Kong and Hang Seng Bank, focusing on agentic AI in finance for customer interactions, insurance claims, payments and onboarding, and on AI giving dynamic oversight of other AI applications' actions.
Bermuda Monetary Authority — The Bermuda Monetary Authority is consulting on The Responsible Use of Artificial Intelligence in Financial Services in Bermuda, AI guidance for financial firms on applying existing obligations to AI they develop, deploy or use. Covering operational resilience, cyber security, transparency, data oversight, risk management and governance, it adds no new requirements or separate AI licensing regime; responses close 30 October.
Jersey Financial Services Commission — Jersey Financial Services Commission (JFSC) folds AI regulation in Jersey's finance sector into duties already binding on firms, bypassing any bespoke regime built solely for artificial intelligence. Its stance remains proportionate and risk-based, stretching current requirements to reach AI in financial services: oversight structures, human review, data stewardship and continuous scrutiny.
Banca d'Italia — Frontier artificial intelligence models bring fresh security threats, says Bank of Italy, urging prompt AI risk management in banking. Lenders must spot weak points and toughen defences, a push linked to the EU Digital Operational Resilience Act, broadening AI risk oversight in financial services under DORA's mandate.
Banque Centrale de Tunisie (BCT) — Generative AI use at the Central Bank of Tunisia now sits under a fresh internal charter, revealed in its 2025 annual report. Several AI-driven data analytics projects run beside it, plus early talks on where the technology heads next. No tools, deadlines or costs were named.
Bangladesh Bank — Bangladesh Bank has barred staff from entering confidential or sensitive banking data into AI tools under any circumstances, a directive dated 28 June 2026 that marks a step in the country's emerging AI governance in banking. Officials must now secure sign-off from the relevant authority before using AI tools in banking use cases such as drafting notes, shaping policy or analysing data.
Financial Services Commission — FSC revised its Financial Sector Artificial Intelligence Guidelines, covering AI governance in financial services, effective 22 June 2026. Seven principles anchor the framework: governance, legality, subsidiarity, reliability, financial stability, good faith, security — subsidiarity keeps executives accountable for algorithm-assisted decisions, not the AI system. FSS issued a companion AI Risk Management Framework covering AI risk management practices for banks; Korea Financial Security Institute released its own AI Security Guide.
National Financial Regulatory Administration — China's National Financial Regulatory Administration issued guiding opinions on 18 June 2026 on AI regulation in banking and insurance, with four principles including that users of an AI system are responsible for it. Institutions must grade AI uses by risk and route high-risk ones through approval; anything touching risk management, underwriting and claims, credit approval, asset valuation, fund transactions or generative AI directly shaping whether a financial contract is agreed counts as high-risk AI. Those need risk-committee sign-off, human oversight at critical points and reporting to the regulator.
Banco de Moçambique — Mozambique AI regulation gains a formal footing at Banco de Moçambique, which has adopted its own Artificial Intelligence Policy for safe, transparent, responsible internal use. Rogério Zandamela linked it to the bank's digital transformation drive and an earlier AI task force. Central bank AI governance is aimed inward here, not at the wider financial sector.
Cyprus Securities and Exchange Commission — The Cyprus Securities and Exchange Commission (CySEC) warned supervised firms in Circular C786 of growing frontier AI cybersecurity risks, as frontier models can locate and exploit software vulnerabilities with unmatched speed and scale. CySEC reminds firms that the Digital Operational Resilience Act demands sound ICT risk management frameworks against AI-enabled cyber threats.
National Bank of Rwanda — Rwanda AI regulation now covers all licensed banks, insurers, payment providers, pension schemes and microfinance institutions under the National Bank of Rwanda's Guidelines No. 52/2026 on Artificial Intelligence Risk Management for Financial Institutions. Firms must set up board-level governance, validate models and keep humans in the loop, with tighter rules where systems flag suspicious transactions or assess borrower risk. This risk management approach reaches further into fraud screening and anti-money laundering monitoring than most comparable frameworks in the region.
Financial Services Agency — Japan's Financial Services Agency and the Bank of Japan jointly ordered banks to tackle AI cybersecurity risks facing financial institutions head-on. Fast-moving frontier AI capabilities, regulators warn, are fuelling attacks on the sector and demand urgent AI risk management now.
Prudential Authority, South African Reserve Bank — South Africa's Prudential Authority told banks to brace for AI cyber risk in banking, under guidance titled Preparing for Frontier AI and AI Accelerated Cyber Risk. Frontier systems let attackers strike almost as soon as a flaw surfaces, collapsing the gap between discovery and breach.
National Bank of Ukraine — Ukraine's central bank, the National Bank of Ukraine, launched the Discussion Paper on the Ethical and Responsible Use of Artificial Intelligence by Participants of the Financial Services Market of Ukraine, opening debate on AI governance in financial services. Priorities named include shielding consumers, managing risk, keeping systems transparent and maintaining operational resilience across the sector. Officials frame this as an early move toward future sectoral guidance and a fuller White Paper, with no binding rules proposed yet under this responsible AI use in banking initiative.
Central Bank of Nigeria — The Central Bank of Nigeria's baseline standards oblige banks, fintechs and payment firms to adopt automated anti-money-laundering systems, the bank's first AI AML rules. They favour automated risk scoring, anomaly detection and behavioural pattern recognition, demand at least yearly independent validation of machine-learning models and require explanations for every alert.
Bank of Ghana — The Bank of Ghana has made AI governance in banking central to its revised Cyber and Information Security Directive (CISD) 2026, introducing AI and Machine Learning Governance in place of the 2018 framework. Regulated lenders must show their AI systems in financial institutions are transparent, fair and secure, covering credit scoring, fraud checks and client support, with boards answerable for cyber risk. A data sovereignty rule requires core systems and customer records to stay within Ghana, applying to payment providers, fintechs, microfinance firms and banks.
Guernsey Financial Services Commission — The Guernsey Financial Services Commission has issued a Policy Statement on the Use of Artificial Intelligence that openly welcomes AI in financial services — machine learning, large language models, plus agentic and generative systems — as a way to run leaner and shed paperwork. Because the Bailiwick supervises by principle, such technology already fits the regime in force: nobody needs prior sign-off or even a conversation with the Commission, and an AI build is treated no differently from any other technical or strategic undertaking under the Finance Sector Code of Corporate Governance and the Minimum Criteria for Licensing. Nothing bespoke accompanies this AI adoption stance — no rulebook, no guidance, no accreditation or certification standard. Where a firm is unsure how to proceed, the pointer is towards the NIST AI Risk Management Framework, ISO/IEC 42001 and the NCSC Guidelines.
Gibraltar Financial Services Commission — Gibraltar's Financial Services Commission has revised its AML/CFT/CPF Guidance Notes to tackle AI risks in AML/CFT/CPF compliance. Three areas gained new provisions: customer risk assessments must retain human sign-off; customer due diligence now flags AI-generated deepfakes; and ongoing monitoring covers AI agents and discriminative or predictive AI models. The changes aim to keep controls resilient as artificial intelligence in financial crime compliance evolves.
Nepal Rastra Bank — Nepal Rastra Bank's draft AI Guideline binds every licensed bank, financial institution and payment provider it oversees. Boards and senior managers answer for AI credit scoring, fraud detection, risk management, compliance monitoring and customer service outcomes, sorted under a tiered risk classification. Clients must be told whenever automated systems shape choices touching them.
Magyar Nemzeti Bank (Hungarian National Bank) — Banking regulators in Hungary have made AI governance in banking mandatory: the Hungarian National Bank (MNB)'s Recommendation 13/2025 (XII.3.) on Digital Transformation of Credit Institutions supersedes the 2021 version and inserts a fresh section on responsible AI deployment. By 30 June 2026, lenders must produce an AI sub-strategy and lodge it with the MNB, meeting standards touching AI risk oversight, data handling, cybersecurity for AI tools, in-house skills and how AI is communicated to the public. Most obligations begin 1 April 2026.
Monetary Authority of Singapore — MAS has proposed Guidelines on AI Risk Management for financial institutions, opening consultation on AI governance across Singapore's financial sector. Expectations span the AI life cycle: data handling, fairness, transparency, human oversight and third-party exposure. Scaled by firm size and reliance on generative AI and AI agents, the rules mean smaller players face lighter duties than larger, heavily automated ones.
Bank of Thailand (BOT) — Thai bank and payment-firm boards now bear direct responsibility for AI outcomes under Bank of Thailand rules built on fairness, ethics, accountability and transparency (FEAT). Its AI Risk Management Guidelines demand human oversight, strong data quality, explainable models and cybersecurity protection at every deployment stage, reinforcing wider AI regulation for financial services.
Central Bank of Jordan — Lenders in Jordan face fresh duties on governance, risk controls and data privacy after the Central Bank of Jordan (CBJ) rolled out an AI regulatory framework for banking. Institutions must keep AI models under constant scrutiny and staff dedicated units with clear accountability. A Steering Committee on Artificial Intelligence and Emerging and Future Technologies takes charge of rollout, tying this to wider AI governance in the banking sector.
National Bank of Georgia — Georgia's central bank has opened its Regulatory Laboratory to fintech firms testing AI in banking, letting companies trial new tools under supervision before wider rollout. The National Bank of Georgia's sandbox for AI-driven financial technology lets advanced systems enter the sector safely, replacing unsupervised deployment. Applications close October 2025.
Otoritas Jasa Keuangan (OJK) — OJK guidance sets minimum standards for AI governance in Indonesian banking, spanning design, testing, rollout, monitoring and audit across the full system lifecycle for banks. Reliability, accountability and human oversight anchor the framework. Non-binding as soft law, it still marks out OJK's supervisory expectations for responsible artificial intelligence deployment in banking.
Financial Services Authority (OJK) — OJK's Buku Panduan Tata Kelola Kecerdasan Artifisial Perbankan Indonesia (Artificial Intelligence Governance for Indonesian Banks) sets out supervisory expectations for commercial lenders adopting AI in banking, published on 29 April 2025 by Otoritas Jasa Keuangan. Institutions are expected to form a dedicated AI committee, apply controls across the full model lifecycle, conduct stress testing and validation, and keep humans in the loop for oversight. The document carries no binding force for AI governance for banks, yet OJK treats it as a reference point when examining banks' risk management practices.
Financial Services Commission, Mauritius — Mauritius's Financial Services Commission (FSC) binds regulated financial institutions to nine principles for AI in financial services. Fintech Series Guidance Notes No. 4 on the Responsible Use of Artificial Intelligence in Financial Services lists fairness, bias mitigation, transparency, accountability, privacy, security, sustainability, human-centricity, monitoring and ethics, effective 19 September 2025. This marks a formal step toward AI governance in insurance and wealth management.
Swiss Financial Market Supervisory Authority FINMA — FINMA's Supervisory Guidance 08/2024 sets expectations for AI governance financial sector oversight in Switzerland, flagging risks institutions must manage: model risk, data quality issues, IT and cyber vulnerabilities, growing third-party reliance, and legal and reputational fallout from artificial intelligence risk management failures. Finding most firms still at an early maturity stage, FINMA now requires processes to identify, assess, control and monitor these exposures.
Securities and Futures Commission — Hong Kong's Securities and Futures Commission (SFC) now holds licensed corporations to conduct standards for generative AI language models used in regulated activities. Senior leadership must answer for outcomes, oversee model risk, and guard against cyber and data threats. Scope covers any firm offering AI-driven services tied to regulated business, effective immediately.
Qatar Central Bank — Binding AI regulation in banking now applies to every QCB-licensed entity in Qatar that builds, buys, outsources or runs AI systems, under the Qatar Central Bank's Artificial Intelligence Guideline. High-risk deployments cannot proceed without prior QCB sign-off, and firms must put board oversight, an AI Register, human review, data safeguards, security measures, and customer transparency and complaint routes in place. Yearly disclosure of AI use to QCB rounds out this AI governance in financial services regime.
Hong Kong Monetary Authority — Hong Kong Monetary Authority guiding principles now cover generative AI in banking, setting GenAI consumer protection safeguards for authorised institutions in customer-facing tools. The principles build on the 2019 BDAI framework, requiring embedded safeguards before deployment. This extends existing Hong Kong Artificial Intelligence Use oversight rather than a separate regime.
Financial Conduct Authority — No dedicated rulebook is coming for AI in UK financial markets: the FCA says existing rules already cover it. Its AI Update points to Consumer Duty and the Senior Managers and Certification Regime as sufficient for risks from AI in financial services, sparing firms extra compliance layers under this approach to AI regulation in financial markets.
Financial Services and the Treasury Bureau — Hong Kong's Government set out an open, prudent stance on AI in financial market activity through its Policy Statement on Responsible Application of Artificial Intelligence in the Financial Market. Financial institutions must balance innovation against risk management, adopting a risk-based AI governance strategy for financial institutions.
Financial Services Authority (OJK) — OJK, Indonesia's financial regulator, released the Guidelines for Responsible and Trustworthy Artificial Intelligence Ethics in the Financial Technology Industry with four fintech associations. Four pillars underpin AI in fintech oversight: beneficial, fair, accountable, transparent, explicable, robust, secure. The non-binding code covers digital lending, payments and insurtech, signalling artificial intelligence oversight in financial technology rather than binding rules.
Financial Supervisory Commission — Taiwan's Financial Supervisory Commission has issued a policy framework for AI regulation in the financial industry, resting on six pillars: accountability, fairness, privacy, system security, transparency and sustainability. It guides how banks and insurers apply artificial intelligence in banking and insurance.
Dubai International Financial Centre Commissioner of Data Protection — Personal data run through autonomous and semi-autonomous systems in the DIFC now falls under Regulation 10, the first AI data processing regulation in the MEASA region. Operators must publish transparency notices, log risk registers, and win Commissioner-authorised certification ahead of any high-risk deployment, while a designated autonomous systems officer handles ongoing oversight.
Bangladesh Bank — Bangladesh Bank's Guidelines to Establish Digital Bank mandate round-the-clock AI-powered complaint handling in digital banking, under Section 9.3. Every licensed digital bank must run a system driven by artificial intelligence that will automatically settle transaction disputes, with no downtime allowed.
Banco Central de la República Argentina (BCRA) — Argentina's central bank, the Banco Central de la República Argentina, requires AI risk management in financial entities under Communication A 7777, folding AI/ML into its security rules via Section 4.6. Institutions must document each algorithm, assign ownership, run impact assessments, and set risk appetite for machine learning deployment in financial services.
Central Bank of Barbados — The Central Bank of Barbados set supervisory expectations for AI risk management in banking through its Technology and Cyber Risk Management Guideline (TCRM), issued 5 May 2023, covering cyber threats tied to automated systems and duties to classify and report cyber incidents. This frames cyber governance for AI-driven banking systems as a regulatory priority.
Securities and Exchange Commission Nigeria — Nigeria's Securities and Exchange Commission now polices algorithm-driven investment advice through its Rules on Robo-Advisory Services, covering how computer-driven advisory platforms register, operate and manage risk. Providers of these automated tools must also satisfy governance standards and safeguards protecting clients.
Financial Services Commission, Mauritius — Mauritius now regulates AI financial advisory services through a licensing regime for robo-advisory services under the Financial Services (Robotic and Artificial Intelligence Enabled Advisory Services) Rules 2021, made by the Financial Services Commission under section 93 of the Financial Services Act. Licensees must hold minimum capital of MUR 600,000, build algorithm governance frameworks, and file independent evaluation reports on their systems every two years.
Tulane University — Tulane University has rolled out TulaneAI, a private generative AI platform offering faculty and staff the latest chatbot models. Hosted on Microsoft Azure, the campus AI service lets users pick models from OpenAI, Anthropic, Meta, DeepSeek or xAI, with no data stored or used for model training.
Angra Pequena Senior Secondary School — Angra Pequena Senior Secondary School in Lüderitz has opened two AI tutoring labs worth N$1.15 million, funded by BW Kudu (N$900,000) and the Namibian Ports Authority (Namport) (N$250,000). YYeni AI's software coaches pupils step by step and doubles as a classroom helper for staff, backed by teacher training and a resident support specialist. YYeni counts over 4,940 tutoring hours and 2,253 practice tests logged in the machine-guided study rooms.
Marietta City Schools — Marietta City Schools in Ohio is piloting MagicSchool AI, a K-12 AI platform, with every educator and student district-wide. Data security drove the choice: MagicSchool is built so user information is never shared or sold. Rather than banning classroom AI, district leaders see the pilot as practical digital literacy training.
Midland Public Schools — Midland Public Schools in Michigan has opened SchoolAI, its AI tool for students and teachers, to every middle and high school pupil after a spring pilot. Engineering teacher Corey Spurling says SchoolAI can create worksheets and translate, and he built a classroom AI assistant pupils can quiz instead of waiting for him.
Ministry of Education, Science, Youth, Culture and Sports of Central Bosnia Canton — Central Bosnia Canton's Ministry of Education, Science, Youth, Culture and Sports will pilot AI teaching assistants through Roboznalac, backed by REZ under a Travnik agreement of 15 September 2026. Funded with 40,000 KM, halved between partners, until 15 December, workshops open the scheme; nine teachers and 30 students in Kiseljak, Travnik and Bugojno then build helpers for three subjects. A digital platform should later give every relevant teacher these classroom AI tools.
University of Santo Tomas — Gemini for Education now backs teaching, research and academic support for over 46,000 students across 22 academic units at the University of Santo Tomas, part of the Philippine institution's push on generative AI in higher education. More than 2,000 faculty gained training to use the tools within their own subjects, with 45 e-learning specialists and over 300 Google Certified Educators providing support, an example of AI in university teaching and learning at scale. Uses span nursing students rehearsing emergencies, architecture students visualising designs, and staff drafting rubrics, lesson plans and visual guides.
Casper College — Casper College has become Wyoming's first college to give all students and faculty campuswide generative AI access, through the BoodleBox AI platform. The rollout began in August 2026 after a select staff group trialled BoodleBox over the summer. Digital Learning director David Siemens links the shared AI workspace to building AI literacy, responsible use and effective teaching integration.
University of Sydney — The University of Sydney will give every student and staff member free, supported ChatGPT Edu from OpenAI, phasing generative AI in higher education into teaching, research and operations. The campus AI tool will sit beside Cogniti. A $1 million seed investment also starts the AI Sustainability Research and Innovation Fund.
Duluth Public Schools — Duluth Public Schools has opened Google Gemini to district staff this school year, putting AI lesson planning in the hands of educators alongside unit development and assessment writing tied to state standards. The choice followed a review of national best practices, principals' association guidance and existing school board policies. Access runs through official district accounts, privacy training comes first, and pupils are locked out; only staff touch the AI planning tools.
Masarykova univerzita — Masarykova univerzita has built AI in higher education administration straight into IS MU, the system its students and staff already use for coursework and teaching logistics, rolling out new capabilities through September 2026. Plain-language queries about study or administrative matters can now be typed directly from the IS MU homepage, answered by a large language model in university information systems that runs on the institution's own servers so personal data never leaves campus. The university calls itself the first in the Czech Republic to embed such a model directly into its academic records platform, and one of the earliest examples anywhere in Europe, with teaching staff expected to shed some routine workload as a result.
Dordt University — Dordt University, a private Christian university in Sioux Center, Iowa, has launched a one-year autumn pilot in which students, staff and faculty get premium versions of AI tools such as Claude, ChatGPT, Gemini and Perplexity via the BoodleBox platform. Monthly campus-wide AI discussions accompany the access.
Government of El Salvador — El Salvador's AI tutoring in schools has already spread past 1,000 campuses, President Nayib Bukele said, pledging coverage of every state school within 18 months. His statement accompanied published findings from a pilot across 171 public schools. Machine tutoring runs beside ordinary classroom teaching. The expansion figures are the government's own.
Ministry of Education, Singapore — Singapore's Ministry of Education has opened Markly, its AI essay marking tool, to secondary and junior-college geography teachers. Its feedback on essays and extended responses is grouped by content and language use, and pilots in some secondary schools cover history, social studies and English language. Teachers review, edit or reject each marking suggestion before students see it and own all feedback; the ministry will judge the pilot before any wider rollout.
SIMAD University — OGAAL SIMAD Chat, from SIMAD University's AI Institute, handles queries on admissions, tuition, courses and services. Built from 63-plus documents plus 1,250 data points, this AI chatbot in higher education serves staff, families and learners bilingually, functioning as a Somalia AI university assistant reachable without campus trips or lines.
Ministry of Education — Trinidad and Tobago's education ministry launched an AI learning platform for secondary students, centred on an interactive AI avatar that walks pupils through problems step by step. Spanish, ICT, integrated science, social sciences, English language arts and mathematics are covered, rolling out term by term, with laptops for Form Two teachers of those subjects. Education minister Michael Dowlath says one more explanation often separates a student who cannot do something from one who understands.
University of Klagenfurt — The University of Klagenfurt is offering every student Academic AI, a free data-protection-compliant AI assistant for students, from the 2026/27 winter semester, for exam preparation, programming, drafting texts and structuring content. The university has no access to students' prompts or chats, and about 30 minutes of basic AI training on Moodle is required before first use.
Armenia Education Initiative — Seroond Authored School began teaching on 1 September 2026 at Vahagni, an AI-native public school the Armenia Education Initiative bills as the country's first, built with the Hovnanian family. Up to 20 children form the opening year group, and nobody pays fees. Machine tutoring pinpoints each pupil's level, tailors drill difficulty live and holds a child back until mastery is real; planning, differentiation and first-pass marking shift to the software, leaving staff for mentoring. AI ethics and AI literacy sit on the timetable.
Kyzylorda Region education authorities — Kyzylorda region has installed an AI system in 144 schools as a pilot of AI for teachers. Some 20,000 staff trained; teachers get a digital helper for lesson materials and tests. Camera analytics alert the director's phone to fights or misconduct in this classroom AI rollout.
University of South Africa — Unisa has rolled out EduBot, an AI chatbot for student support that answers queries typed in a student's own words and preferred language. Unresolved cases pass to a live agent who sees the prior conversation. Unisa calls it a complement to existing support services, an AI-powered online assistant rather than a replacement.
TI-Hüpe (AI Leap Foundation) — TI-Hüpe, the AI Leap scheme, now reaches every gymnasium year, basic-school teachers, and vocational teachers and students, spreading AI in Estonia's classrooms. At Tallinna Pelgulinna Riigigümnaasium, pupils already work with the programme's AI learning application.
Kwame Nkrumah University of Science and Technology — A tie-up between Kwame Nkrumah University of Science and Technology and Google for Education has extended AI tools in higher education to 95,000 people, with Google Workspace for Education Fundamentals licences reaching students, staff and administrators. Recipients gain entry to Gemini and Gemini Notebook, forming a Gemini AI rollout for university communities. Online preparation for lecturers launched September 2026, with a Gemini Academy following in November.
University of Copenhagen — The University of Copenhagen earmarked DKK 110 million for a three-year programme from 1 September 2026 bringing a generative university AI assistant, the 'KU AI-assistant', and AI transcription to administration, teaching and research. Shared AI infrastructure and three research AI labs follow, led by Morten Axel Pedersen, its first vice-rector for AI.
Houston ISD + 54 US school systems — Houston Independent School District and 54 other US school systems have adopted OpenAI's ChatGPT for Teachers for lesson planning, communications and admin. OpenAI says that adds nearly 140,000 educators and staff, taking its free programme past 340,000 users in 103 partner organisations across 30 states. Houston ISD's innovation director Christine Mai says teachers decide when to use AI and check its accuracy.
Wamego USD 320 — Wamego USD 320 in Kansas has deployed ZeroEyes' AI gun detection and situational awareness platform in its schools, the company says. It works with the district's existing security cameras to flag images that may show a firearm; possible detections go to ZeroEyes' operations centre for round-the-clock review by staff from military or law-enforcement backgrounds. The district (just under 1,600 students across four schools) paid via the Kansas Safe and Secure Firearm Detection Grant Program, run by the state attorney general's office.
Peninsula School District — Peninsula School District is giving every staff member personal, agentic AI assistants that can handle multi-step tasks on their own. Chief information officer Kris Hagel said teachers are building interactive learning games, science simulations and flight simulations for its aeronautical academy, with the aim that staff have more time to be human.
Columbus County Schools — Columbus County Schools has adopted the MagicSchool AI platform district-wide for 2026-27, giving teachers and students district-controlled, education-focused AI tools. The district intends it to help teachers build lessons and support students' writing and maths within teacher-monitored guardrails.
Pender County Schools — Pender County Schools, which introduced MagicSchool AI for staff last year, is extending it to students this term, with teachers designing assignments that build in supervised student use. The district picked the platform for its monitoring tools and student-safety guardrails, and adopted a Generative AI Guidance Policy (Policy 3221) alongside updated technology and discipline policies.
Saint Xavier University — Saint Xavier University is giving faculty, staff and students BoodleBox, a collaborative AI platform brokering models such as GPT, Claude, Gemini and Perplexity, for 2026-27. Staff and faculty trained in spring 2026; the university calls itself Chicago's first with campus-wide access. Students can cite and share AI-assisted work, letting professors grade the process as well as the product. Provost Carees Dean linked it to readying students for academic success, career and life.
Ministry of Education of Kazakhstan — Kazakhstan's education ministry, under Minister Zhuldyz Suleimenova, said its AI in schools pilot would begin on 1 September in ten small schools across the Pavlodar and Kyzylorda regions, with AI solutions tested in fourth-grade classes covering mathematics, digital literacy, the native language and AI itself. Fifty more schools are scheduled to join in November and another 300 in January 2027, on the way to 500 small rural schools through 2029. Published a fortnight before the start date, the source does not confirm that AI-assisted learning began on the announced day.
MINESEC — MINESEC assembled over 250 bilingual AI lab simulations in under a month, free and runnable offline, unveiling the secondary-school programme at the Centre AIDL on 17 August 2026. Muluh Godson, National Pedagogical Inspector for Computer Science, gave those figures; Chemistry, Biology, Arts and Letters and Industrial Techniques staff ran demonstrations, and Inspector General Adjaba Biwoli Jean Pierre cast the purpose as learning made interactive, practical, reachable. Minister Nalova Lyonga cautioned that AI stays a tool, pressing structuring, reasoned use and prompting, and asking teachers to raise pupils as creators. Cameroon aims these AI-generated science experiments at schools holding no physical laboratory.
University of Colorado — The University of Colorado opened ChatGPT Edu to students under a system-wide agreement with OpenAI, but each must first finish a 30-minute online module on AI privacy, data protection and appropriate use. Boulder's Office of Information Technology announced the student rollout after faculty and staff got access earlier this year, and instructors still set AI rules for each course.
Ain Shams University — Roughly 10,000 first- and second-year students at Ain Shams University now use "Learning Buddy", built by the Faculty of Commerce with Microsoft, Telecom Egypt and EduSofx. It is Egypt's first institution-wide AI learning platform, generating coursework, adapting study routes per learner, and flagging pupils falling behind for staff, beyond the earlier AI-driven educational platform trial on a marketing course.
Duke University — Duke University is offering a suite of AI tools for autumn 2026, free for students, staff and faculty with a NetID: ChatGPT for Education, Microsoft Copilot, its own DukeGPT platform and a MyGPT assistant builder, with paid options including Microsoft 365 Copilot and, later in term, Anthropic's Claude. NetID sign-in keeps work within Duke's protected environment, the Office of Information Technology says.
Charleston County School District — Charleston County School District has started a one-year pilot of MagicSchool, an AI tool for teachers' lesson planning and classroom preparation, with wider rollout expected later in autumn 2026. Teachers district-wide attended summer generative AI training with the AI for Education Foundation, and the district's AI framework leaves classroom use up to teachers.
National University of Singapore — The National University of Singapore is giving all students, faculty and staff ChatGPT Edu from 31 August 2026 under an OpenAI collaboration. It also launched AI Sense Maker, a conversational AI research platform built on 150,000-plus digitised NUS Libraries materials for topic summaries, semantic search and knowledge-graph creation. Professor Aaron Thean says the tools amplify, rather than replace, human intellect.
University of Johannesburg — University of Johannesburg has equipped MoUJi, its multilingual AI chatbot, with large language model capability for isiZulu, Afrikaans, Sesotho and English conversations. Users can sort admission checks, Admission Point Score sums, class sign-ups and financial aid enquiries whenever suits them, in their chosen tongue. UJ links the AI-powered chatbot to lighter contact-centre load and fewer temporary registration-period hires, an unverified institutional claim.
Direction Départementale d'Éducation du Nord-Est, MENFP — Haiti's North-East departmental education directorate now applies AI exam quality control before state examination results are finally validated, a step in modernising public education administration. Later that month the education ministry released first results for the 9th fundamental year, the baccalaureate and the ENIJE teacher-training exam.
University of Verona — The University of Verona is trialling ChatGPT Edu with professors and young researchers ahead of AI-literacy courses. Those courses on AI skills start next academic year.
American International University West Africa — Gambia now hosts a female AI-powered robot for STEM education, built by American International University West Africa for schools short on science and maths staff. Dr Shukla, university president, called it a machine that answers instantly and behaves naturally with those it meets, an AI robot in science and maths classrooms rather than a stand-in for scarce teachers. Maker Labs supplied the software and AI layer; engineers in Bangalore, India built the physical frame.
UCT Online High School — UCT Online High School has launched Msizi, an AI in education assistant answering routine questions from parents and learners so staff can tackle harder cases. It scans pupil results to build resources like study guides for 4,000-plus registered pupils. Academic head Tessa Venter described the tool as boosting teachers within remote schooling, not replacing them.
Papua New Guinea Department of Education — Papua New Guinea's Department of Education unveiled EduNex, an AI-powered learning management system, at Port Moresby National School of Excellence. The AI learning platform, built on homegrown NEXA AI, gives teachers assessment tools, digital classrooms, curriculum resources and AI-assisted lesson planning. Education Minister Francis Siune expects it to connect schools nationwide; Starlink and DataCo supply internet first to National Schools of Excellence.
Ministry of Education of Belarus — Belarus grades six to nine now cover AI in schools under Ministry of Education resolution No. 132, dated 22 July 2026, effective 1 September 2026. Grade eight gets a unit explaining how AI systems in the classroom function and guiding pupils toward responsible, appropriate use, across 139 weekly hours total.
Western Sydney University — Western Sydney University has made Microsoft 365 Copilot available to all staff, casuals included, and Higher Degree Research candidates from 16 July 2026, bringing generative AI in higher education to presentations, meeting summaries and document drafting. After a 100-user pilot, vice-chancellor George Williams tied the move to student success, and the university claims to be first in New South Wales and second in Australia to give all staff enterprise-grade Copilot.
Ministry of Higher Education and Scientific Research (Algeria) — Algeria's Ministry of Higher Education and Scientific Research has opened an AI university orientation assistant for 2026 baccalaureate holders, alongside a digital edition of its ministerial orientation circular. According to the ministry's notice, the orientation chatbot replies instantly on every step of choosing a course and enrolling.
Stanford University — Three AI assistants reached every Stanford University account holder at once: OpenAI ChatGPT Edu, Google Gemini Enterprise and Anthropic Claude for Education, part of a higher education AI adoption pilot lasting through August 2027. Gemini Enterprise turns on by itself for anyone holding a Stanford Google login, whereas the other two demand single sign-on plus sign-off on responsible-use rules first. Buying pressure from research teams and staff already paying out of pocket for these tools pushed Stanford, in the United States, toward this campus-wide AI tool licensing, which beats individual subscriptions on both data protection and cost.
MTN South Sudan, Ministry of General Education and Instruction, Educare Organization South Sudan — MTN South Sudan has launched Junub Academy, a national e-learning platform providing AI-supported education in South Sudan free of data charges. Partners are the Ministry of General Education and Instruction, the Ministry of ICT and Postal Services, and Educare Organization. Students and teachers gain curriculum materials, teacher training, digital skills courses and AI-driven learning tools, backing the government's Free Education Policy.
Ministry of Education, Science, Culture and Sports — Armenia's Ministry of Education, Science, Culture and Sports says it will hand roughly 50,000 free ChatGPT Edu seats to the country's educational, engineering and research communities when the school year opens, under an arrangement with OpenAI and Firebird. Universities will open the platform to lecturers and undergraduates, while AI in schools reaches every teacher and senior pupil in the state system. Staff are expected to draft assignments, tests and topic plans with the software, freeing hours for inventive work, with briefing and groundwork due by September. Firebird covers the bill through 2026-2027, and ministers will weigh how classroom AI performs before deciding whether the treasury takes over.
Ministry of National Education and New Citizenship — Snote Manager, swapped in by the Democratic Republic of Congo's Ministry of National Education and New Citizenship for the retired Data Scan, brings AI-assisted exam marking to the 2026 Examen d'Etat. At 10 to 15 times its forerunner's speed in pulling in answer sheets, the package embeds automated script processing. Capture sites multiplied from one to six, so papers are read close to candidates then moved under encryption onto central servers rather than hauled overland to Kinshasa. Sheet capture opened 22 June; Inspector General Hubert Kimbonza counted Haut-Katanga 1 and further educational provinces cleared.
Ceibal and ANEP — Ceibal and education authority ANEP have added Gemini for Education (Plus version) and NotebookLM to the AI tools for teachers already in CREA and the Matific and Aleks maths platforms. Teachers may request data-protected institutional accounts, use is voluntary, and Ceibal's president promises monitoring and evaluation of the effects.
Ministry of National Education (Türkiye) — Türkiye's Ministry of National Education launched DİLİM, an AI foreign-language learning platform, on 16 June and later reported 220,417 users. The platform analyses speech and writing live with personalised feedback, while a teacher module shows pupils' progress alongside class and individual performance reports.
Jigawa State Government — Jigawa State will offer AI home learning to state-school pupils who miss class through farm labour, poor weather or other interruptions, via the JigawaUNITE programme with NewGlobe. Officials say the AI-powered education system will also coach teachers in class, guide instruction live and track performance.
Ministry of Education Malaysia — The Education Ministry made SMK Kuala Perlis the first school nationwide to adopt the AI-Powered Classroom initiative, bringing AI in schools under the Pre-Google Reference Institution programme. Built with Google Malaysia, the pilot sets the template for a school AI rollout: 27 further schools next phase, 800 within seven years.
Ministry of Education (Israel) — Israel's Education Ministry will move every middle-school pupil to AI English learning next school year, with teachers guiding each pupil's progress. The ministry also plans to widen Project 720, which pairs pupils with an AI learning assistant and shows teachers live where pupils struggle, from about 28 pilot schools to 180 middle schools.
Governo de Mato Grosso do Sul (Secretaria de Estado de Educação) — Mato Grosso do Sul's state government has signed a Google partnership bringing AI in state schools to about 190,000 students through Gemini and Google Workspace for Education, with training for up to 5,000 teachers. The state education secretary says pupils and teachers will need guidance on appropriate, ethical use of Google's AI tools.
International Professional University of Technology — Proof of skill with Google's Gemini chatbot, through a fresh Google Gemini certification for AI in higher education, will reach every learner and employee at International Professional University of Technology, spanning its Tokyo, Osaka and Nagoya sites. Underpinned by Google AI Pro for Education, the plan mixes web-based lessons with formal testing so graduates can later point to generative AI skills certification when job-hunting.
Education Development Fund, Prime Minister's Office (Egypt) — Egypt's Education Development Fund, part of the Prime Minister's Office, introduced an AI ADHD assessment through the Egyptian Center for Testing on 15 May 2026. The AI-scored test returns instant, plain-language results so parents and teachers can grasp a child's attention, focus and hyperactivity patterns and make sounder schooling choices.
Ethiopian Ministry of Education — Ethiopia's Ministry of Education and Huawei began trialling a national-curriculum AI teacher assistant for grades 9 to 12 on 14 May 2026. Working entirely without internet access, the classroom AI lets teachers draft exams and lesson plans faster where networks are weak.
TI-Hüpe, Institute of the Estonian Language — Iti is the winning name for the school AI tutor fielded by TI-Hüpe and the Institute of the Estonian Language, chosen with 198 of 442 ballots ahead of Tõps on 85, Klaabu on 61, Säga on 50 and Raju on 48. Over 11,000 teenagers in grades 10 and 11 at 128 upper secondary schools already work with software shaped with OpenAI and Estonian academics. Vocational classrooms and thousands more sixth-formers join the AI learning application next academic year.
Ministry of National Education, Preschool and Sports (Morocco) — Morocco's education minister, Mohamed Saad Berrada, told parliament AI exam marking checks will join the baccalaureate's review and audit stages alongside human markers. The aim is fewer errors and automatic re-verification of results.
RF Merchant Bank & Trust (Barbados) Limited / WizdomCRM — RF Merchant Bank & Trust (Barbados) Limited has widened AI financial literacy at Harrison College, whose 300-plus entrants push the Sustainable Stock Market Artificial Intelligence Game to almost 800 pupils with The Lodge School and Combermere School. Caribbean tech firm WizdomCRM delivers the game, backed by the Ministry of Education Transformation. Virtual portfolios tracking domestic and overseas exchanges cover financial planning, market trends and investing via artificial intelligence-powered learning tools, AI-assisted investing lessons widening on 8 May 2026.
Ministry of National Education, Literacy, Technical Education and Vocational Training of Guinea — Guinea's Ministry of National Education will launch an AI exam-prep tool called Barrin to support baccalaureate candidates preparing for their exams. Minister Alpha Bacar Barry revealed the AI-powered study companion at a STEM teacher training session, saying it would go live the following week. Students will gain access to educational resources, including worked solutions drawn from earlier mathematics papers and other subjects.
CUBE The Gambia — CUBE The Gambia is piloting its home-built CUBE AI Tutor on 150 secondary pupils across three Region 1 schools, trialling AI tutoring in Gambian classrooms: St Augustine's Senior Secondary School, St Joseph's Senior Secondary School and Charles Jow Memorial Academy. The offline-ready, curriculum-matched tool, an AI learning platform for secondary students, adds speech-to-text and text-to-speech. Pupils trial it for two months, then give feedback.
Ministry of Science and Education of the Republic of Azerbaijan — Azerbaijan's Ministry of Science and Education has wired a bespoke Copilot Agent, spun up in Microsoft Copilot Studio, into the Digital School platform, so AI lesson planning sits in the app teachers open daily. The helper roughs out plans and classroom handouts, boils down background reading, frames slide decks and turns out first-cut assessments and rubrics. Some 900,000 accounts sit there: 44,552 teachers, 543,255 pupils, 360,709 parents.
Institute of Public Administration and Management (IPAM) — IPAM, Sierra Leone's Institute of Public Administration and Management, built Academic Hub, an AI essay marking and online exam platform, to boost integrity and transparency in higher education. Trials with students went well, so officials plan wider rollout, including at the University of Sierra Leone. The AI-supported online examination system should improve grading accuracy and cut result delays.
Ministry of Education, Trinidad and Tobago — Trinidad and Tobago's education ministry launched AI digital textbooks in English and mathematics for Form One students on 21 April, on laptops issued after the Secondary Entrance Assessment. Pupils can quiz an AI avatar about schoolwork during and after school, and education minister Michael Dowlath expects savings of TT$1,000 to TT$1,500 for Form One and Two parents next year.
Ministry of Education and Science — Latvia's Ministry of Education and Science launched a national-scale AI in education programme for general and vocational schooling, backed by a €200,000 Google.org grant. Plans include an AI leader school network of 15 general and 5 vocational schools, new teaching materials and a revised national curriculum. By 2030, students, teachers and school leaders should use educational AI skilfully and responsibly.
An-Najah National University — An-Najah National University has introduced an AI virtual patient, developed by the digital transformation unit of its media centre, for sixth-year medical students' final examinations at An-Najah Hospital. The interactive avatar converses by voice and video without disclosing the condition, forcing students to reason towards a diagnosis, and scores each performance to give examiners data-driven assessments. A case library spanning specialties is planned.
State of Thuringia (Ministry of Education) — Thuringia has made telli, an AI chatbot for teachers, available to state and independent schools through its school portal; pupils get no direct access. Education minister Christian Tischner called it a historic step.
Nihon University — Ten thousand full-time employees at Nihon University gained access when Google AI Pro for Education launched campuswide, marking Japan's single largest paid generative AI deployment university administrators have undertaken. Staff and faculty work inside a closed, secure platform that excludes their data from training the underlying models. Supporting this shift, a newly formed Information Innovation Centre will push the school's broader AI in higher education strategy forward.
MIRAIE — Future Education (MIRAIE) built FUTURE COMPASS, a generative AI platform for high-school inquiry-based learning, on Google's latest AI. It coaches students on sharpening research questions and action plans, giving teachers live visibility into group thinking. This AI-assisted inquiry-based learning tool pilots at model schools from Hokkaido to Okinawa.
Instituto Superior de la Judicatura de Panamá (ISJUP), Órgano Judicial — Panama's judicial training institute, the Instituto Superior de la Judicatura (ISJUP), launched ISJUP IA, an AI study assistant for judicial staff and legal professionals' ongoing training. Curriculum designers, teachers and students can use the virtual assistant to consult course content, settle legal questions and find study materials.
Akwa Ibom State Ministry of Education — Akwa Ibom State launched an AI teaching programme in public primary schools with EIDU and the state Ministry of Education, offering pupils personalised learning and teachers structured lessons. Phase one targets about 6,000 pupils in 67 schools in 2025/26, and live data lets teachers and heads spot learning gaps early.
State of Lower Saxony (public schools), via FWU — Lower Saxony has made telli, an AI chatbot for schools commissioned by all 16 German states, free for every public school, says FWU, the states' media institute. Teachers and pupils get data-protection-compliant generative AI, with the language models hosted only on EU servers.
Sir Arthur Lewis Community College — Sir Arthur Lewis Community College (SALCC) debuted Arthur, an AI chatbot for student applications, for the 2026/2027 intake. Trained from SALCC's site and FAQs on OpenAI-compatible technology, it answers WhatsApp queries on programmes, entry requirements and fees. This AI admissions assistant cuts staff's routine manual replies.
Mentora — Tangier now hosts Mentora, Morocco's first AI-powered educational platform, covering learners from primary grades through the baccalaureate. Over 50 specialists spent six years building it around Ministry of National Education material, adding adaptive quizzes and study routes in Arabic, French, English and Moroccan Arabic. It signals a push for curriculum-locked AI tutoring software matched to exam requirements.
Ministry of Education (Kuwait) — Kuwait's Ministry of Education is rolling out Google for Education's AI tools to about 40,000 secondary teachers and department heads, an AI teacher training project for safe, ethical classroom AI use. Running from early 2026 to year-end across all educational districts, it seeks common digital proficiency standards and better teaching practice.
Ministry of Digital Transformation of Ukraine — Ukraine's Mriya education platform now offers an AI test generator for teachers, automating assignment creation and marking to curb cheating. It draws on 1,300-plus digitised textbooks across 30 subjects and ran in 150 schools, producing over 4,000 assignments. The Ministry of Digital Transformation aims to widen this AI-based assessment tool to 500 more schools before a nationwide launch.
Mauritius Ministry of Education — Mauritius's Ministry of Education and Mauritius Telecom introduced mytGPT Education on 26 January 2026, a generative AI teaching assistant trialled with pupils in Grades 4, 7, 8 and 9. The classroom AI generates quizzes and lesson resources and shows teachers student performance, in English, French and phonetic Creole, following the national curriculum.
Ogun State Government — Ogun State Governor Dapo Abiodun launched OgunLearn, an AI learning platform for the state's schools offering students lesson resources and interactive support. The AI-powered teaching tool should help teachers prepare, deliver and monitor performance more efficiently, within an education digitisation agenda alongside the Diper school management system.
Comunidad de Madrid (Regional Ministry of Digitalisation) — Spain's Community of Madrid says it will add AI tools for teachers to its EducaMadrid schools platform: first an email assistant that drafts, reviews and polishes, then one designing personalised pupil content automatically. Five schools will trial the features first; each teacher could save about 40 routine hours a year, it estimates.
NTU Singapore — NTU Singapore has given all undergraduates a Google AI suite (Vertex AI, Google AI Studio, Gemini Enterprise) plus its own NTU AI Learning Assistant (NALA). President Ho Teck Hua aims to embed AI in higher education across 40% of courses in 52 undergraduate degree programmes by 2030, from about 5% now, with students building and deploying AI agents on real problems. NTU calls itself Singapore's first university adopting AI at this scale.
The US Department of Labor is building AI skills into Registered Apprenticeship instead of creating stand-alone AI occupations, and launched an AI in Registered Apprenticeship Innovation Portal in April 2026. On 7 July 2026 it announced $162 million in performance-based incentives, $40 million of it to Jobs for the Future for AI, semiconductor and nuclear-infrastructure apprenticeships. It served over 58,000 registered apprentices in technology, cybersecurity and AI-related occupations in 2025.
Singapore's Ministry of Manpower says Singaporeans taking selected government-funded SkillsFuture AI courses will get six months' free access to premium AI tools from the second half of 2026. Pairing subsidised training with subsidised tools is meant to get learners applying skills through regular use and experimentation; qualifying courses and tools will be named later, according to its March 2026 press release.
Estonia's national AI Leap programme for AI in schools expands to vocational schools from September 2026, adding 38,000 students and 2,000 teachers across vocational schools and the new grade-10 intake after a first phase of 20,000 upper-secondary students and 3,000 teachers from September 2025. The government is negotiating with AI developers including Anthropic and OpenAI to supply tools through a public-private AI Leap Foundation, an initiative of the President's Digital Council and the Ministry of Education and Research.
Ontario Tech University — Ontario Tech University has put an in-house AI Learning Agent into 22 undergraduate and graduate courses across every faculty for the Winter and Spring 2026 terms. Drawing only on instructor-approved course materials, it prompts students with guiding questions instead of giving answers, and faculty control access and content. Instructors see recurring student questions live, helping them target teaching, the university says; in this pilot, student testers help refine and govern the tool.
Moroccan Ministry of National Education, Pre-School and Sports — Morocco's Ministry of National Education will use AI exam cheating detection devices to find and block suspicious communications in baccalaureate examination centres for the session opening in early June 2026. Named T3-SHIELD and made by Sensthings, a company linked to the Mohammed VI Polytechnic University, the signal detectors catch Bluetooth, Wi-Fi and cellular transmissions and other hidden communications, and serve no other purpose. Education-policy academic El Houssine Zahidi warns technology alone falls short without family awareness work.
ESBanque (École supérieure de la banque) — French banking school ESBanque has introduced Mon Coach IA, AI role-play training built with Sopra Steria and Gwenlake, where future bank and wealth-management advisers rehearse client meetings with AI-driven virtual clients. Trialled with 90 learners since June, it was set for over 1,000 from September 2026.
Facultad de Ciencias Económicas, Universidad de Buenos Aires — The University of Buenos Aires' Faculty of Economic Sciences approved AI rules for teaching, learning and assessment and began running its own AI teaching assistant inside the virtual classroom, loaded with each chair's reading list. The in-house assistant already serves 40 subjects across 184 courses, about 2,000 students.
Vancouver School Board — Microsoft Copilot chatbot accounts reached secondary pupils aged 13 and up across Vancouver School Board, Canada, part of a wider push on AI in schools. Guardrails filter harmful material, shield privacy and prevent chats from training AI models, while staff describe the effort as building safe, thoughtful habits among learners. New teacher guidance arrived alongside this classroom AI rollout.
Chatmore International School — Chatmore International School has switched Years 2 to 11 onto CENTURY's AI learning platform, a Bermuda first, following a pilot. Staff sought bespoke study routes, instant feedback and usable data without extra workload. Secondary teacher Sandrine McGlashin records no pupil objections, with the AI-driven teaching tool mostly set for homework and brief lessons; head of school Britanni Butterworth points to stronger attainment among IGCSE pupils wanting extra drilling. More than 70 COBIS schools worldwide run CENTURY.
Elizabeth College — Elizabeth College has run a month-long pilot of Inkling, an AI tutor built by UK independent school Clifton College, which recruited the Guernsey side to test and feed back. Instead of supplying solutions outright, this AI-guided study aid applies the Socratic method — probing, never telling — to build critical thinking, and logs where each pupil stands against exam objectives so help arrives tailored. Fran de Garis, English teacher and AI lead, put Year 10 through it for homework beyond class time, with cover lessons a further possibility. De Garis stops short of committing: adoption stays undecided, the hoped-for role an extra option rather than a crutch, and results go to Inkling's developers in September.
Energiaügyi Minisztérium — Hungarian vocational schools have opened a 10-hour course, 'A mesterséges intelligencia alapjai', midway through this academic year, and AI teaching in vocational education must reach every ninth-grader by 31 December 2026. A trial reached 1,700 learners at 63 sites, 60 home-based and three abroad, says the Energiaügyi Minisztérium; Logiscool supplies syllabus, platform and staff training. School AI lessons should leave pupils able to wield it, follow its mechanics and spot dangers.
FHI 360 — FHI 360 piloted an AI chatbot for teacher training with 16 primary school instructors and instructional leaders in Madagascar's Atsimo-Atsinanana region under its FIANTSO project, a three-week trial in January. This AI teacher-training assistant answers pedagogy questions in Malagasy, including regional dialects, using Ministry of Education-validated materials. Feedback was strong: 70% rated it very useful, and 80% said it was simple to use.
Jeonjin Advanced Middle School — Jeonjin Advanced Middle School, Pyongyang's first integrated K-12 institution, has built a dedicated AI education room in Nangrang District with educational robots and up-to-date teaching gear. The facility delivers staged AI in K-12 education: kindergarteners engage with robot play, elementary pupils tackle simple algorithms, intermediates move into programming, and advanced learners design and operate robots themselves. This graduated model of AI-based robotics teaching reflects a push to embed artificial intelligence instruction at every school stage.
Ministry of National Education of Senegal — Senegal's Ministry of National Education has rolled out Allô Gorgui, an AI assistant for students and teachers, developed with local startup Andakia and unveiled at the Concours général 2026 awards ceremony. It is an AI-powered educational assistant that gives spoken lesson explanations and exercise help in French, with Wolof support being added gradually, over an ordinary phone line with no smartphone or internet connection required. Coverage is set to widen to further national languages later on.
Ministry of Education and Human Resource Development — Seychelles' Ministry of Education and Human Resource Development and the World Bank are running two pilots of AI in schools. The AI Teacher Coach Programme helps Primary 5 and Primary 6 teachers at six schools plan lessons and grow professionally. The AI Student Tutor Programme gives Secondary 3 learners at four schools personalised classroom AI support in Mathematics and Geography.
Ministry of Basic and Senior Secondary Education (MBSSE) — Sierra Leone's Ministry of Basic and Senior Secondary Education finished EDITH (Education Digital Information Tool Helper), an AI in education management platform turning days or weeks of administrative work into minutes. Development followed selection for the EdTech Hub Ministerial AI Challenge. The Ministry's IT department is set to take over the AI-assisted education administration tool permanently.
Succeed Individual Tutoring Institute — Succeed Co now runs School AI, an AI tutoring platform, across its Individual Tutoring Institute Succeed chain. Sister firm Mingaku built the system; over 1,600 schools nationwide already use it so staff can draft lesson plans while pupils get an AI study companion for unsupervised revision.
Ministry of Digital Transformation of Ukraine — AI Mentor education platform is now live on Diia.Education, giving Ukraine's 3 million-plus users a Gemini-based virtual helper backed by Google.org. It simplifies jargon, condenses course content and produces quick summaries. An AI Guide navigator comes next, the platform's first step into broader AI-driven features.
Taoyuan City Government and Wenxin Elementary School — Generative AI now reaches classroom coursework at Wenxin Elementary in Taoyuan, where Taiwan's first AI smart classroom launched on 23 December 2025 through a joint effort by the Taoyuan City Government, ASUS and Google for Education. Named the AI Chain-of-Thought Smart Classroom, the space allows pupils to draw on such tools for assignments and gives instructors support in building materials and interpreting learning analytics. Mayor Chang San-cheng signalled the approach might spread to further schools should outcomes turn out well, marking an early instance of AI in classroom education in Taiwan.
Mohammed VI University of Health Sciences (UM6SS) — Mohammed VI Foundation for Sciences and Health launched Inteligis UM6SS, giving the Mohammed VI University of Health Sciences community access to AI in health education. It offers self-paced modules on artificial intelligence in medical science, joint research spaces, and tools that automate tasks, personalise study plans and produce videos.
University of Surrey — Full staff access to My AI Surrey has arrived at the University of Surrey, an AI platform in higher education blending Cloudforce's nebulaONE technology with Microsoft Azure hosting. Semester two, February 2026, brings AI tools for staff and students to every learner, unlocking coursework guidance, research aid and hands-off paperwork processing.
University of Nottingham Malaysia — Salesforce's Agentforce platform powers NOVA (Nottingham Online Virtual Assistant), University of Nottingham Malaysia's new AI assistant for student applications, answering queries on eligibility, scholarships, housing and booking appointments across the website, WhatsApp, Facebook and WeChat. UNM is Malaysia's first university to deploy this AI virtual assistant in higher education.
Ministry of Communications and Information Technology (MCIT), Ministry of Education and Higher Education (MOEHE) — A new platform for AI in public schools, the National Personalised Learning Platform, was unveiled at World Summit AI Qatar 2025 by Qatar's Ministry of Communications and Information Technology alongside the Ministry of Education and Higher Education, developed with Scale AI. Student records will be examined to shape lessons around each child's pace and strengths, and educators gain automated support for marking work, drafting plans and monitoring progress over time. Sitting within the wider Digital Agenda 2030 strategy, the effort marks a push toward personalised education technology reaching every state-run school nationwide.
Université de Tunis El Manar — University of Tunis El Manar has rolled out a digital platform, at www.puit-utm.com, applying artificial intelligence in higher education to pair students with firms for placements and roles. Matching weighs skills against labour-market demand, while employers gain visibility into the graduate talent pool via AI-driven job matching.
Ministry of Education — 20,368 Grades 10 and 11 learners have joined Guyana's biggest online school under an AI-driven personalised learning in schools initiative, launched 5 December 2025. Educators guide live digital classes backed by a broad e-library, while artificial intelligence shapes lessons to fit each student. Reach now stretches past Guyana into several CARICOM neighbours.
Senegal's Ministry of Education and Kajou — Tutobac IA, a free AI Baccalaureate exam prep tool, now runs inside Senegal's Kajou app for Android and iOS. It generates tailored explanations and practice questions from the official Bac syllabus, an AI-assisted Bac revision aid before the national test. Lycée d'Excellence Mahmady trialled it 4 December 2025 ahead of nationwide rollout.
Mohammed V University — Mohammed V University in Rabat built ELOGHA-SUP, Morocco's first homegrown AI language-learning platform, spanning five languages across six clusters nationwide. Chatbot-driven AI tools power the interactive study experience. It marks a milestone for domestic AI-driven language education software.
Ministry of Education and Training (Vietnam) — Grades 1 through 12 across Vietnam will trial an AI teaching in schools scheme run by the Ministry of Education and Training, lasting through May 2026 at chosen sites. Four strands anchor the syllabus: human-centred thinking, AI ethics, AI techniques and applications, and AI system design. A second track brings artificial intelligence in education directly to staff, equipping them with tools for planning and grading; further details remain pending.
Nepal Academy of Science and Technology (NAST) — 42 science teachers from Kathmandu's Chandragiri and Tokha municipalities gained AI tools for teachers instruction under Nepal Academy of Science and Technology, mastering ChatGPT, Gemini and DeepSeek. Practised tasks include exam-paper setting, report drafting and material design. Rollout of AI in teacher training to further municipalities is planned.
Universidad Autónoma del Perú — Peru's Universidad Autónoma del Perú has introduced URUK, an AI-powered virtual learning platform built and trained entirely in-house, without relying on an external vendor. Fed by the institution's own library holdings, the system delivers automatic summaries, indexes texts, answers queries instantly and converts readings into podcasts for study, applying the Cornell note-taking method. This puts AI-assisted studying tools at the centre of the university's academic strategy.
I. Horbachevsky Ternopil National Medical University — I. Horbachevsky Ternopil National Medical University has deployed an AI virtual patient for medical education, Ukraine's first, built with the Institute of Cybernetics of the National Academy of Sciences. This AI-based virtual patient tool for clinical training tests trainees on history-taking and communication skills across paediatric, surgical and gynaecological exam stations for 2025-2026.
Universidad de las Américas Puebla (UDLAP) — Universidad de las Américas Puebla uses predictive AI to flag students at risk of dropping out, running on Microsoft Dynamics 365 and covering more than 35 thousand student, alumni and candidate records. It identifies those who miss re-enrolment deadlines and automatically triggers tailored outreach to re-engage them while they are still deciding whether to stay or leave. The tool is part of a wider digital transformation at the Mexican university, applying student attrition prediction before learners finalise that decision.
Government of Rwanda (Ministries of Education and ICT & Innovation) with Anthropic and ALX — Chidi, an AI learning companion for schools built on Anthropic's Claude, now reaches Rwandan classrooms via the Ministries of Education and ICT & Innovation. Pupils get coding, data science and problem-solving help; staff use the same AI-powered education tool for lessons. Up to 2,000 teachers and civil servants get AI training too.
Cavendish Education — Five UK schools run by Cavendish Education, serving neurodivergent students, now trial an AI-powered teaching tool for SEN pupils reaching about 250 learners. Otermans Institute built the system around a digital avatar that converses two ways, answers pupil questions instantly and gives personalised feedback. Staff keep authority over what lessons cover, watching each child's progress through an AI teaching assistant for neurodivergent students.
Fundación Emprender Futuro — Free WhatsApp coaching for small business owners in Bolivia now runs through Arien, an AI assistant for entrepreneurs from Fundación Emprender Futuro and partner Banco de Crédito de Bolivia (BCP). Debuting at Global Entrepreneurship Week (GEW 2025), it skips separate downloads and walks users through ChatGPT prompt techniques, cryptocurrency basics, QR payment setup, catalogue design and money-saving habits. Workers and founders alike can tap this AI-powered digital skills chatbot to build stronger tech and money management abilities.
Lyuben Karavelov Primary School, Vidin — Lyuben Karavelov Primary School in Vidin now delivers AI personalised learning, running a system that builds lesson outlines in seconds once subject, difficulty and outcomes are set. Pupils get AI-driven individualised instruction matched to their pace, backed by new science labs and an education museum.
University of Agder (UiA) — Norway's University of Agder now uses an AI grading feedback tool to draft mark explanations, checked before reaching students. It fields 10-12,000 such requests yearly; 67 per cent of staff polled backed the approach. Built since 2023, the system covered 36 modules in spring 2025, aiming for 100 modules of automated grade explanations by autumn.
Ministry of Education and Children / Directorate of Education and School Services — 600 teachers across Iceland now hold access to Anthropic's Claude and Google's Gemini, under a scheme run by the Ministry of Education and Children to bring AI in education into everyday classroom work. The Directorate of Education and School Services oversees delivery with the Icelandic Teachers' Union, covering lesson prep, resource building and paperwork. Spanning October 2025 to April 2026, findings will guide broader AI-assisted teaching rules nationwide.
Odense Municipality — Odense Municipality gives every pupil and teacher access to SkoleGPT, an AI tool for schools built for Danish classrooms. Data stays protected during coursework and research, and staff get training on using it in lessons — a full AI tool for pupils rollout, not a trial.
ABA Agency — A homegrown artificial intelligence model now powers EduPro, marking the first AI in education platform built entirely in Palestine and across the wider Middle East. ABA Agency designed this AI tutor platform to reach six groups — pupils from grade 4 through high school, parents, teachers, schools, directorates and the Ministry of Education — via a virtual AI tutor plus adaptive smart exams, behavioural analytics and dashboards suited to each user. The Higher Council for Creativity and Excellence has backed the launch, hosted on private Palestinian servers for full data sovereignty.
Ministry of Education (Kenya) — A new AI school placement pilot from Kenya's Ministry of Education is sorting 1.13 million Grade 9 learners into senior secondary schools. Basic Education Principal Secretary Julius Bitok said the software weighs each pupil's grades, local capacity and chosen career track, seeking fairer geographic spread as the Competency-Based Education reforms take hold. The trial launches with the first group shifting up in January 2026, marking an early test of automated secondary school allocation nationwide.
Rising Academies — An AI maths tutor called Rori now reaches Grade 7 pupils at Madiana Basic Cycle School via WhatsApp, under a six-month Rising Academies scheme. The Ministry of Education's Regional 2 West Education Directorate co-runs it, sending each child individualised numeracy exercises by phone message.
Baden-Württemberg Ministry of Education — Baden-Württemberg's education ministry is bringing AI in schools statewide, giving every school access to telli, an AI chatbot built into the SCHULE@BW digital platform. The classroom AI assistant offers conversational practice, tutoring partners, and scenario-based exercises for lesson prep, with privacy safeguards officials say will improve through ongoing teacher and student feedback.
Primary School Svetozar Markovic (Kragujevac) — Smartgorithm's Cvrle, an AI chatbot in primary education, made history at Kragujevac's Primary School Svetozar Markovic, Serbia's first trial. Second-graders met this conversational AI in schools during Children's Week for a civics session on friendship, norms and rights via dialogue, building digital literacy.
Federal Directorate of Education — One urban and one rural school in Pakistan have begun trialling grade 8 lessons built around AI in education, through a scheme run by the Federal Directorate of Education alongside technology firm Mind Hive. Machine assistance now supplements standard curriculum delivery, replacing parts of conventional teaching. A positive outcome could justify wider rollout, eventually covering up to 100,000 students.
Ministerium für Schule und Bildung NRW — KIMADU, a North Rhine-Westphalia pilot, brings AI agents in education to 25 secondary schools, producing custom maths and German study paths. Staff authored the megaprompts, calibrated per year group and skill level, drawing on curricula and digital textbooks. State backing beyond one million euros funds the scheme, among Germany's bigger classroom AI deployments.
NSW Department of Education — NSW Government is expanding NSWEduChat to public school students in Years 5 to 12 from Term 4, after a 50-school AI in schools pilot. Built in-house on the Department of Education's own cloud environment, the generative AI app steers pupils toward answers via guided questions instead of handing them over, fostering critical thinking.
University of Oxford — Oxford is the first UK university giving all students and staff ChatGPT Edu access, capping a year-long, 750-person trial. A five-year OpenAI deal, struck March 2025, backs this ChatGPT Edu higher education push, part of a wider AI in higher education shift.
Corporación Universitaria Minuto de Dios (UNIMINUTO) — UNIMINUTO now runs two generative AI agents for AI in higher education across Colombia, reaching over 42,000 students and a thousand-plus teachers. MÍA mentors on emotional, academic and personal matters; LEO, an AI teaching assistant in 349 courses, reportedly cuts lesson-planning effort up to 70 per cent, unverified.
Tilburg University — Tilburg University's developers coded Tilly in-house, an education AI chatbot letting lecturers fill it with their own course materials for exam rehearsal or subject-specific queries at any hour. Hosted on Microsoft Azure via European data centres, this AI tutor platform avoided any outside vendor entirely.
Sechenov University (First Moscow State Medical University) — An AI assistant for medical students called DocAI, built inside Sechenov University's digital department, began trial use in autumn 2025. Positioned as an AI medical education tool, it pulls together clinical literature, tunes explanations to what each student already knows, and connects lessons to clinical guidelines. Rollout continues to Samara State Medical University by spring 2026, then across every medical school in Russia.
Peking University — Guanghua School of Management and Peking University's Computing Centre built "Doujiao," an AI teaching assistant for humanities and social-science instruction: knowledge-base assembly, outline and PPT drafting, live quiz interaction, and natural-language lookup in the school's case library. Staff compile reference sets, sketch lecture materials, run polling, and pull cases without formal query syntax. The rollout marks a first push into generative tools for liberal-arts subjects, distinct from science and engineering AI in higher education uses.
VŠB-Technical University of Ostrava — 60 computer science students at VŠB-Technical University of Ostrava tested a round-the-clock, multilingual AI teaching assistant during the Unconventional Algorithms and Computing course, reaching it through a web interface to fix coding bugs, work through hard algorithms and check materials. Having proven itself, the scheme now spreads into cybersecurity and quantum computing modules, broadening AI-assisted learning across Czechia's higher education.
UNSW Sydney — UNSW Sydney will hand 10,000 staff access to ChatGPT Edu in higher education, calling it Australia's largest education deal with OpenAI. A 10-month pilot with 800 staff preceded the move, with 98% asking to stay on the AI tool for university employees. Coverage extends to all fixed-term and permanent staff.
Ânima Educação — Ânima Educação, Brazil's largest higher-education group, has launched IA.RA, an AI learning assistant on its Ulife platform. Built for AI in higher education, it produces summaries, questions and reading suggestions, checks assessments, and translates and transcribes Portuguese, English and Spanish. The AI-based teaching assistant personalises study and cuts workload.
Gobierno de Venezuela — Schools and high schools across Venezuela will teach AI classes in schools from the 2025-2026 academic year, under a directive from President Nicolás Maduro. The plan links this new AI education in Venezuela to building homegrown, sovereign technology skills, preparing a future workforce of engineers and programmers.
South Australian Department for Education — SA will extend EdChat, a ChatGPT-style AI chatbot in schools built with Microsoft, to every South Australian public high school next term, says Education Minister Blair Boyer. The classroom AI assistant was trialled with educators and 10,000 students.
Sapienza Università di Roma — Sapienza University of Rome has enrolled 50 students in Italy's first randomised clinical trial of AI in medical education. Four oncology and precision medicine modules pair virtual tutors with simulated patients, offering tailored feedback and interim quizzes against a conventionally taught control group. This AI-based teaching method's value will be judged on clinical skill gains, satisfaction and diagnostic error rates.
Sindh School Education and Literacy Department — 3,500 educators across Dadu, Tando Allahyar, Tharparkar and Umerkot are set to receive AI teacher training via Pakistan's Sindh School Education and Literacy Department, working alongside UNICEF and Khan Academy Pakistan. Across six months, staff will practise using Khanmigo, an AI assistant for lesson plans, materials and pupil engagement, blending live instruction with independent study modules. Authorities call it an unprecedented AI-based teacher training pilot nationally.
Daystar University — Daystar University now runs ArthurAI, MindHYVE.ai's agentic AI in higher education platform, across campus under a new partnership. It flags at-risk students via predictive analytics, adapts lesson pathways per learner, works offline on mobile, and delivers content in over 10 languages. East Africa gains one of its first campus-wide agentic AI learning platform deployments, launched for the September 2025 semester.
Ministry of Education of Greece — Greece's government partnered with OpenAI, extending AI in secondary education to 20 Greek Lyceums via Onassis Foundation backing. December 2025 brings ChatGPT Edu access for staff, then pupils by March 2026 under this generative AI in public schools initiative. Such national-scale classroom rollout remains rare worldwide.
University of Europe for Applied Sciences Dubai — UE Dubai gives every Digital Business and Data Science student Buddiz AI, a personalised AI assistant for higher education. Trained on course syllabi, assignments and exam requirements, it adapts through use as an academic AI assistant for students built into the curriculum to sharpen critical thinking about AI-generated output.
Centre de services scolaire des Chênes — Centre de services scolaire des Chênes in Drummondville added four generative AI tools — Microsoft Copilot Chat, Canva, Google Gemini and Gemini Notebook — to its toolkit. Secondary, vocational and adult-education students, teachers and special-education technicians got access first, ahead of administrative and professional staff from 15 September and primary pupils later, extending AI in Quebec schools via generative AI in education.
SESI-SP — LEIA, an AI tutor in schools from SESI-SP on Microsoft Azure OpenAI, covers all 140 São Paulo state schools, roughly 36 mil students and 3 million-plus monthly interactions. It reshapes material around each learner's speed, interests and struggles beyond class time. Staff mine usage patterns to catch gaps and rework teaching, an AI-powered learning platform now central to instruction.
AI Council under Ministry of Industry and New Technologies — Tajikistan's Ministry of Industry and New Technologies has begun offering AI education in schools through more than 100 institutions nationwide, using its AI Council to deliver a free weekly period to grade 10 and 11 pupils. Dushanbe hosts sixty of the sites, while 40 operate in outlying towns and districts. Since 1 September 2025, this optional scheme has served as groundwork for eventually making 'Fundamentals of AI' a mandatory curriculum topic, with AI lessons in schools not yet compulsory.
Niigata University — ChatGPT Edu, OpenAI's education-grade generative AI platform, went live for Niigata University staff and learners on 1 September 2025, making it among the earliest Japanese national universities to roll out generative AI in higher education. Managed, secure access now underpins a newly launched practical generative AI course plus trial schemes spanning teaching, learner support and administrative work. The rollout builds on a course recommendation system driven by AI that the university launched in May 2025, after discovering that roughly 60% of incoming students had never used generative AI tools during secondary school.
Museo Nacional de Historia Natural (MNHN) — The Museo Nacional de Historia Natural (MNHN) in La Paz, Bolivia, premiered Biopochito AI with 60 students from Colegio Ruso Boliviano, who used the AI science education app to spot biodegradable waste, assemble robots out of reused scrap and study how nature is protected. Built by the International Green Science Center for Latin America and the Caribbean (IGSCLAC) with the Andean Road Countries for Science and Technology (ARCST), this AI-powered science learning tool reframes difficult ecological and scientific ideas as play-based games and tasks for children and young people.
Indiana University — Indiana University is deploying ChatGPT Edu to its entire 120,000-strong community, OpenAI's second-largest ChatGPT Edu higher education rollout. Staff gained access September 2, 2025; students follow January 1, 2026. A 200-instructor trial found 80% said the AI chatbot for universities aided teaching, research and service work, and a new free course, GenAI 101, now builds responsible AI habits campuswide.
Saint Clare School for Special Education — AI in special education cuts planning time at Saint Clare School for Special Education by 30 to 50 per cent, via a Learning Management System built with SOZCODE. Using over 15 years of data, it turns basic learning needs into matching goals and activities from a stored SPED bank for staff to adjust, speeding Individualised Education Plans and special needs teaching.
Corporación Unificada Nacional de Educación Superior (CUN) — CUN, with Google, is Colombia's first university centring AI in higher education on teaching and administration, part of wider AI adoption in higher education. Gemini, NotebookLM, Google Classroom and AppSheet span every course for a student body 80 per cent virtual.
Tottori Prefectural Board of Education — 365 students at Tottori model schools now use School AI, a generative AI platform for education, following its rollout by the Tottori Prefectural Board of Education on 20 August 2025. Under a Ministry of Education scheme for AI-enhanced English education, the system sharpens speaking and writing through conversation drills, feedback on essays and vocabulary matched to each learner. Staff also receive tools for setting tasks and marking work, a rare instance of AI in English language teaching led by a regional authority.
Miami-Dade County Public Schools — Miami-Dade County Public Schools rolled out Google's Gemini for Education to about 100,000 high-schoolers, cementing its rank as the third-largest US district embracing AI in K-12 education. Learners lean on the assistant to shape ideas and get instant translation for English-language learners, a case of generative AI classroom tools reaching an entire student body.
Maxis (with Ministry of Education Malaysia) — 150 schools, 30,000 people: Maxis's target for AI training for students and teachers, backed by Malaysia's Ministry of Education. Under Program GenAI Maxis eKelas, tracks cover prompt engineering, ethical use and real-world applications. Early trials of this generative AI workshop for schools gave a 40% average lift in post-test scores.
Harbin Institute of Technology — HIT's Faculty of Computing engineered 'Gong Xiaoxing,' a 24-hour AI counsellor for students at Harbin Institute of Technology, fusing empathetic dialogue with tiered logic: menus for routine questions, warm replies for distress, retrieval-augmented generation for thorny procedures — delivering AI mental health support in higher education. Over 2,000 students and counsellors shaped it, covering study advice, admin queries, and stress relief.
Shandong Jianzhu University — Shandong Jianzhu University is now Shandong province's first with a full-stack AI education platform, unveiling Zhisiti, an education large language model (LLM) spanning 70-plus subjects. Assistants field queries non-stop across 24 hours, generate lesson plans automatically, break down multi-modal lectures and flag pupil conduct.
Universidad Nacional Mayor de San Marcos (UNMSM) — Universidad Nacional Mayor de San Marcos built BEA, an AI harassment support tool for spotting warning signs, learning rights and finding reporting channels. MediaLab, its research group, coded the assistant on ChatGPT, distributed via OpenAI's GPT Store. Short for Bienestar Estudiantil contra el Hostigamiento, BEA is among Peru's earliest campus uses of generative AI against harassment.
Tunghai University — Taiwan's first AI PC classroom, an AI computing lab for education, opens at Tunghai University: 75 ASUS PCs, donated by an alumna, powered by Phison aiDAPTIV+ cards. NTD 100 million funds campus AI infrastructure, including an NVIDIA GB200 server, for training large language models and generative AI on site.
Newark Public Schools — Newark Public Schools will deploy Amira Learning's AI literacy screener across all K-3 classrooms from autumn 2025, under a board-approved contract capped at $900,000 over two years. The tool listens to pupils reading aloud, scoring accuracy, pace and expression to flag those at risk of falling behind, functioning as an automated reading diagnostics system. Adoption forms part of New Jersey's revamped literacy framework for early reading.
Ceibal — Ceibal built AI lesson planning into its CREA learning platform, giving Uruguay's teachers a secure space to draft class materials with machine help. The feature, PowerBuddy, sits inside CREA and suggests activities and educational resources through AI-assisted content generation on request.
University of Insubria — A 24/7 stand-in covers Italy's University of Insubria: the country's first digital twin professor, cloning lecturer Davide Tosi into a 3D avatar with the University of Bologna. Fed his classes, slides and text, it speaks in Tosi's voice, sparing hesitant learners a direct approach via an AI virtual teaching assistant.
Universitat Politècnica de València — Universitat Politècnica de València built PoliGPT, an AI chatbot for higher education, using only internal staff. Live since spring 2024, it answers campus queries and handles subtitling, transcription and translation for UPV Media and VideoApuntes, plus natural-language search. This reflects rising university AI adoption in Spain via homegrown systems.
KAIST — KAIST runs a retrieval-augmented Virtual Teaching Assistant (VTA) for 477 students on its Programming for Artificial Intelligence course, an AI teaching assistant mining lecture slides, assignments and recorded lectures for 24/7, evidence-based answers. Over half the cohort used it regularly, racking up nearly 3,870 Q&A exchanges, while questions reaching human TAs fell roughly 40% year on year. KAIST has open-sourced the code, extending this virtual tutoring model beyond its own campus.
Qatar Foundation Pre-University Education (QF-PUE) — Qatar Foundation (QF) has deployed Google Gemini for Education across its Pre-University Education schools with partner Advanced Business Computing. Unveiled at the Google Cloud Summit 2025 in Doha, the AI in K-12 education push targets better teaching and pupil engagement via personalised tools. Generative AI in schools now spans classroom and admin functions at QF.
Beeline Kazakhstan — Beeline Kazakhstan's Janymda super-app now hosts an AI language tutor, engineered by QazCode atop KazLLM, a homegrown large language model. Six playful modules with a guiding avatar serve grown-ups and youngsters studying Kazakh. Rarely does such an AI-powered learning assistant rest on domestic rather than imported technology.
Centre de services scolaire De La Jonquière — Marking time per text has dropped from 15 to 5 minutes since Centre de services scolaire De La Jonquière began trialling AI essay marking via Emilia, a Quebec-built correction software, in Saguenay–Lac-Saint-Jean classrooms. The tool also pinpoints each learner's weak spots, guiding teachers toward tailored follow-up. Trials are expanding into Mauricie as Emilia's rollout progresses.
Universal AI University — Karjat's Universal AI University has launched GURUJI, an AI chatbot that instantly resolves academic and administrative questions, an early example of AI in higher education. Running on machine learning and natural language processing, it recaps missed lectures and draws context from PDFs, websites and other linked sources. This AI chatbot for students works as a round-the-clock campus companion.
Vinschool — Vinschool, Vietnam's largest private system, runs an AI curriculum kindergarten to grade 12 at 54 campuses, ages 5-18. VinUni University guided it, folding AI into IT and computer science rather than standalone classes, covering mechanics, literacy and responsible use via this AI education programme for students, from 2025-2026.
Turks and Caicos Islands Ministry of Education — 1,000 pupils in Turks and Caicos Islands get free access to the Eagle Eye AI Tutor and Sustainable Stock Market Game AI Platform, part of AI tutoring for secondary school students. WizdomCRM Caribbean Limited built the six-month programme, covering money handling, investing and saving through an artificial intelligence financial literacy programme. TCI is the fifth Caribbean territory to join.
University of Graz — Graz is Austria's first university with an AI chatbot for students, giving all 30,000 enrolled free access to studiGPT. Learners use this AI study assistant to shape schedules, sharpen writing, craft messages and spark ideas. The rollout paired with a new micro-degree, AI and Society, carrying 16 ECTS credits.
EtonHouse Education Group — Teachers at EtonHouse now spend one to two hours on lesson preparation instead of a full day, after the Singapore provider rolled out Lumina, an AI lesson planner powered by OpenAI's application programming interface. Administrative workload has dropped 90 per cent since the generative AI system replaced an older Excel-based method. Piloted at selected centres first, Lumina tracks learning objectives as they happen, works across multiple languages, and keeps a familiar look borrowed from conventional planning templates — part of a broader shift toward AI in education.
Tokyo Metropolitan Board of Education — Toritsu AI, a generative AI platform for schools, now reaches 256 metropolitan schools under the Tokyo Metropolitan Board of Education after two years of trials at 29 research schools. Its Tokyo-only tenant excludes prompts from model training, filters unsuitable material, and gives teachers ready lesson templates — the fruit of that pilot phase, now a full AI in education deployment.
Chiba Institute of Technology — Chiba Institute of Technology now runs Japan's first AI university lecturer, built from ChatGPT joined with blockchain-based Verifiable Credentials for AI teaching in higher education. It studies each learner's history and answers through dialogue, tested in the school's web3 and AI course until July 2025. Activity becomes a tamper-proof credential, proving learners' progress.
Quebec Ministry of Education — 22 Quebec school service centres now fully run an AI dropout prediction tool, 27 more mid-rollout. GRICS built the system for the Ministry of Education, scanning grades and attendance to flag Grade 6 through Secondary V students at risk of leaving school early. Mila, Obvia and CTREQ shaped its design, part of predicting student dropout risk.
Ministry of Education, Culture, Science and Technology — Belize's Ministry of Education, Culture, Science and Technology unveiled three AI teaching assistants on the 501 Academy portal at the country's first National EdTech Summit. Avatar Buddy built the set: 501 Historian, covering Belizean heritage material; 501 Teacher Assistant, for lesson planning; and Kriol 501 Historian, replying in Kriol and English. The AI buddies for education target teachers' paperwork and lesson quality.
Ministry of Higher Education, Research, Science and Technology (MoHERST) — Gambia's Ministry of Higher Education, Research, Science and Technology (MoHERST) rolled out TITA, an AI timetable app for Gambian universities that builds class schedules automatically. Nigerian firm TITA Universal Services Ltd built the mobile platform, which also handles course registration, tuition payments, CGPA tracking and assignment submissions at public universities nationwide, centralising tasks once split across separate manual systems via an automatic timetable generator.
Hong Kong University of Science and Technology (HKUST) — HKGAI V1, Hong Kong's first homegrown large language model, is free for HKUST staff and students, reflecting growing AI adoption in higher education and AI use in universities. Built on DeepSeek and tuned for Cantonese, Putonghua and English, it joins 10-plus tools on the GenAI Platform, including GPT-4, DALL-E and DeepSeek-R1, with faculty training via the Center for Education Innovation.
The Ladies' College — A custom-built agent now guides Year 7 pupils at Ladies' College in Guernsey through setting their own goals, an early example of AI in education taking hold in the classroom. Staff at the school rely on a locked-down Microsoft Copilot for planning lessons. Daniele Harford-Fox pointed to this shift in explaining why the curriculum has been redesigned around AI's expanding influence on pupils' prospects.
Ministry of Education (Kuwait) — "Ma3 Hamad Chat" gives Kuwait's Ministry of Education an AI chatbot for education, built in-house to field questions across the school community via app and website. Textbook-grounded replies span several languages and dialects, checked by a review team beforehand. This AI virtual assistant for schools speeds communication and boosts learning effectiveness.
Ministry of Education, Skills, Youth and Information — Several Jamaican schools are testing AI exam marking under a new Ministry of Education, Skills, Youth and Information pilot. Dana Morris Dixon said staff now watch scores appear instantly as software grades scripts, a change designed to hand teachers back hours for direct student contact rather than manual scoring.
Anambra State Government — Anambra State, Nigeria, deployed EdTech platform Edves in 22 pilot smart schools, one per local government area. Adaptive lesson planning, predictive analytics and AI personalised learning shape instruction and monitor student progress live. Teachers get AI-curated lesson templates, performance analytics and automated assessment tools; the AI-tailored learning platform handles enrolment, digital attendance and infrastructure monitoring.
New Windsor School — New Windsor School in Auckland runs an AI literacy pilot built into daily lessons through the Ako Hiko Cluster. Students compose music with Suno, generate video through Sora, and get instant writing feedback from Brisk, sharpening reading, writing and creative skills. Principal Lisa Harland argues outright bans achieve nothing, so pupils learn instead to judge when AI in schools helps and when it hinders.
Macquarie University — Macquarie University trialled Virtual Peer, an AI chatbot for students, across a group of 1,400 undergraduates and recorded nearly five extra marks on average, a 9.45% lift, among those who used it, an AI-powered virtual student assistant that proved its worth in testing. Running on Microsoft Azure OpenAI Service, it handles academic and administrative questions 24/7, drawing solely on verified institutional records via retrieval-augmented generation rather than searching the open web. Satisfaction reached a Net Promoter Score of 92, a striking result for campus-based software of this sort.
Prins Wilhelmgymnasiet, Flen — Judy, an AI teaching assistant shaped as an avatar, guided pupils at Prins Wilhelmgymnasiet in Flen through a civics discussion on human rights, marking its debut leading a class rather than supporting from the side. Staff at the Swedish school oversaw the exchange, having trained on the AI avatar in the classroom during the prior autumn as part of a wider effort to bring artificial intelligence into everyday lessons.
Bavarian universities via bidt and TUM — OneTutor, an AI tutoring system built at the Technical University of Munich, now runs on nine Bavarian campuses this term. Learners get replies pulled from course material, while staff design quizzes. AIffectiveness in Education, run by the Bavarian Research Institute for Digital Transformation, tracks the tool's effectiveness.
University of Vienna — University of Vienna's u:ai gives all staff ChatGPT-based tools for chatbot queries, document analysis, drafting, translation and coding help, an AI assistant for higher education within a data-protection-compliant setup excluding submitted content from training. Austrian universities jointly run the AI platform for university staff to share costs.
Insuco Tiburcio Saavedra / Municipalidad de Temuco — Chile now has its first public liceo running an AI security system: Insuco Tiburcio Saavedra in Temuco, fitted with a metal-detector portal and facial-recognition cameras. Dahua Technology Chile built the setup for the Municipalidad de Temuco after knife-related scares on site pushed officials to act. Trial length: six months, finished late February 2025.
Universidad Iberoamericana (Unibe) — Universidad Iberoamericana (Unibe) deployed an AI virtual assistant for university students, built with CardNET and named Aurora, to handle admissions, enrolment, payments and academic documentation in the Dominican Republic. It answers multilingual queries from students, graduates, staff and faculty via WhatsApp and email, using generative AI with institutional data for real-time processing — an instance of AI in higher education administration.
Shikho — Bangladesh-based edtech company Shikho unveiled a beta AI tutoring assistant for Bangla-speaking students, running on generative models fine-tuned for Bangla-language processing to narrow academic gaps. This AI-powered doubt-solving platform for students serves paying subscribers first, ahead of rollout to teachers, parents and schools. A freemium option follows, designed for cheap smartphones and minimal data use.
King Fahd University of Petroleum and Minerals — King Fahd University of Petroleum and Minerals becomes Saudi Arabia's pioneer in adopting ChatGPT Edu higher education tools, an OpenAI product for learning. Faculty, researchers and learners gain access to the AI platform for universities, aiding scholarship, coursework and cooperative projects. The step advances KFUPM's AI+X programme, obligating each learner to master core AI before picking a specialism.
Khon Kaen University — Khon Kaen University rolled out KKU IntelliSphere, a free multi-AI platform giving Thai staff and students access to nine large language models: OpenAI's GPT-4o, Anthropic's Claude, Google's Gemini, Meta's Llama and others. It is Thailand's first deployment of this scale. Access began January 2025, extending to everyone from 11 February 2025, with over 1,000 already registered — proof of demand for this multi-model AI access.
Israel Ministry of Education — Roughly 500 kindergartens across Israel now form a testbed for Israel's Education Ministry, piloting AI robotics in early childhood education. Youngsters assemble tracks for small robots, crack puzzles as teams and perform robotic theatre, absorbing basic coding along the way. Educators, meanwhile, gain AI tools to craft personalised activities matched to what each child needs.
IU Internationale Hochschule — IU Internationale Hochschule, Germany's largest private university, has added AI agent functions in higher education to its learning assistant Syntea: the system now picks teaching methods on its own and nudges students about their goals instead of only answering questions. Trials began in early 2025 and reached online courses by the first quarter, letting Syntea suggest study approaches, map out personalised paths and fold in tasks such as scheduling courses. IU calls itself among the first in Germany to bring this kind of agentic AI tutoring into the classroom.
Universidad Estatal a Distancia (UNED) — Costa Rica gained an AI assistant for higher education when UNED (Universidad Estatal a Distancia) unveiled Unedia, a metahuman avatar helper — the first at a public university nationally. Generative AI drives the lifelike figure, fielding enrolment, academic and service questions nonstop, 24 hours daily, by text or voice — a university virtual assistant Leandro Camacho built free in two months.
Chulalongkorn University — Gemini 1.5 Flash and Pro reach the whole Chulalongkorn University community through ChulaGENIE, a generative AI platform for universities hosted on Google Cloud's Vertex AI, with Anthropic's Claude and Meta's Llama models slated to arrive later. Staff gained entry first, in January 2025, then students in March. Chulalongkorn pairs this with Google AI Essentials courses, building AI literacy in higher education campus-wide.
AI4EDU project (University of Cyprus and Cyprus Pedagogical Institute) — AI4EDU, a joint project of the University of Cyprus and the Cyprus Pedagogical Institute, put AI classroom tools for secondary education through four classroom trials, dividing 160 teenagers evenly between test and comparison groups of 80 apiece. Physics, Chemistry and English Literature lessons hosted the conversational assistants TeacherMate and Study Buddy, with the latter also backing homework completion and exam readiness beyond ordinary lesson hours, functioning as AI tutoring assistants in schools. The work ran January to May 2025.
Ministry of Education (Oman) — 4,000 Microsoft 365 Copilot licences are heading into Oman's public schools, a Ministry of Education deal with Microsoft ranking the sultanate among the first in the Middle Eastern region rolling out AI in education nationally. Staff get lesson-planning, marking and content help; pupils get tailored support; leaders get performance data. Revealed at COMEX 2025 — part of a wider AI copilot tools in schools push.
Ministry of Education (Ghana) — Over 1.4 million students in Ghana now have AI learning apps for senior high school, an AI-powered SHS curriculum tools push by the Ministry of Education reaching 68,000-plus teachers under the new curriculum. Built from local content, the apps work offline. A prior test phase covered 712 schools before nationwide launch.
Mount Kenya University — Predictive analytics now flag at-risk students at Mount Kenya University under UnIRP, a new resource-planning system built with Marwadi Technologies to embed AI in university administration. Biometric logins secure access, a mobile app tracks student progress, and admission-to-graduation tasks run automatically. A compulsory AI course now awaits every incoming student, part of Kenya's wider AI education push.
Ministry of Education, Innovation, Youth, Sports, Arts and Culture — Classroom instruction at Lüderitz Secondary School now blends with software under a five-month scheme called the YYeni AI Lüderitz Pilot, built on the YYeni AI learning platform and connected via Telecom Namibia. Staff have shifted toward guiding pupils individually as the system adapts material for them across subjects, an AI-assisted learning approach officials hope proves itself. Namibia's Ministry of Education intends to widen this AI tutoring effort to cover 25 schools and 2,500 learners come 2026.
University of Amsterdam — University of Amsterdam's AI chat platform university tool, UvA AI Chat, runs solely on in-house servers, keeping data off outside systems. Staff build subject-specific bots from course content; learners get free access with no premium tiers. The setup favours digital autonomy over commercial AI chat tool for education alternatives.
Tio Business School — Coding knowledge is no longer required for Tio Business School teachers to design their own AI teaching assistants, thanks to a new rollout of Cogniti AI. Twenty instructors have already finished the initial round, and a further 30 are lined up before term restarts in September. Behind the system sits the University of Sydney, which developed the platform and licenses it to Tio, with Blis Digital handling implementation for this push into AI in higher education teaching.
Puerto Rico Department of Education — Puerto Rico's Department of Education handed Microsoft 365 Copilot to 6,000 teachers and administrative staff in a 2025 rollout, feeding a broader push for AI in education, and 15,000 licences now run across the system. Staff lean on the tool, an AI teaching assistant, to draft lesson plans, build assessments, write grading rubrics and adjust materials for neurodivergent learners. Sessions to train personnel carry on into year-end.
MTN Uganda — More than 213,000 mobile money agents and small merchants across Uganda now access an AI chatbot for agent training built by MTN Uganda into WhatsApp. Called MoMo Coach, it pushes short lessons on launching a venture, handling finances and expanding trade, addressing gaps in this AI-driven agent training at scale.
Chiang Mai University — Chiang Mai University in Thailand launched Matthew, a generative AI platform for higher education and an AI-powered academic research tool, built on GPT-4o and Amazon Web Services. Chatbots manage courses, generate and analyse images via DALL-E 3, and search literature through Semantic Scholar. A trial with 100+ lecturers and 4,000 students ran from September 2024 ahead of faculty rollout; students join in 2025's first semester.
National University of Science and Technology — Oman's Ministry of Transport, Communications and Information Technology and the National University of Science and Technology have launched the A2M Platform (BIMANIC), applying AI in higher education to predict student outcomes and guide academic planning and decision-making. The university AI tool was among several 2025 ministry pilots under the National Program for Artificial Intelligence and Advanced Technologies.
Ministry of Basic and Senior Secondary Education, Sierra Leone — Sierra Leone's Ministry of Basic and Senior Secondary Education runs an AI chatbot for education data on WhatsApp for its eight-person delivery team, via the EdTech Hub Ministry of Education AI Challenge. Embedded expert Mohamed Alpha created the retrieval-augmented generation tool, which supplies pupil and teacher numbers or exam results from checked files when staff compile budgets and reports. Endorsed by 17 district directors, this one-year-licence education data assistant gets near-daily use.
Libraries Without Borders (Bibliothèques Sans Frontières) — Libraries Without Borders tests an AI language app with 30 primary school teachers in Dakar and Thiès until March 2025, partnering Senegal's Ministry of Education, with French Development Agency backing. Kajou and Pleias research lab built the AI French tutor on a Llama 7-billion-parameter model, writing tailored drills plus thorough one-to-one feedback.
Independent Institute of Education (IIE) — South Africa's Independent Institute of Education (IIE), its largest private higher education provider, has fitted an AI-enhanced learning platform, D2L Brightspace, across campuses serving nearly 65,000 students. D2L Lumi charts individualised study routes, while Creator+ generates interactive material, giving staff room to scale personal attention.
India's Ministry of Skill Development and Entrepreneurship (MSDE) launched the Skill India Assistant on 21 July 2025, a multilingual AI skilling chatbot on WhatsApp built on Meta's open-source Llama model. MSDE calls it the first large-scale open-source AI deployment in a national public skilling mission. By February 2026 its Skill India Digital Hub, run with the National Skill Development Corporation (NSDC), counted 1.5 crore (about 15 million) candidates, 7,000 training providers and 68,000 employer and industry partners.
A World Bank randomised trial in Edo State, Nigeria showed AI tutoring lifting learning by roughly 0.31 standard deviations, and 0.23 in English, across a six-week after-school scheme for about 800 senior-secondary students. Supervised by teachers, pupils used a generative-AI tutor, Microsoft Copilot powered by GPT-4; the World Bank likens the improvement to about two years of typical learning, and end-of-year exam scores rose too. A formal working paper records the trial.
Ministerio de Educación, Ciencia y Tecnología de El Salvador — El Salvador's education minister, Karla Trigueros, says 171 pilot public schools outscored the national average in science, mathematics and reading, performing on a par with Germany and Sweden, crediting AI tutoring in schools, teacher training, learning monitoring and technology. OECD's PISA for Schools assessed 1,198 students in June 2026; the schools, selected for internet access and infrastructure instead of past results, do not represent the national system, whose PISA 2025 results differed. Extension to all public schools within 18 months is planned, though no independent evaluation has isolated the AI tutors' effect.
Turkish high-school maths students given unguarded GPT-4 scored about 17% worse on later AI-free exams than peers without AI, in a randomised controlled trial of AI in maths education with about 1,000 students (Bastani et al., PNAS, 2025). A guardrailed 'tutor' version that withheld answers erased the damage, making this the strongest causal proof so far that unscaffolded AI use can degrade learning.
A 2026 peer-reviewed evaluation in the International Journal for Educational Integrity found AI-text detectors rarely gave false positives but mostly missed AI-generated and 'humanised' text, with only one tool reliable. The authors say detectors cannot alone establish academic misconduct; false negatives, not false accusations, are the real risk.
An Ofsted review of early AI adopters among English schools and further education colleges found scant independent academic evidence that AI lifts educational outcomes, and no clear guidance on gauging that impact. Most providers tracked staff usage or feedback, not data measuring effects on learners.
A co-designed generative AI teaching assistant at a Singapore vocational institution improved teaching processes but produced no significant gain in student assessment outcomes, an evaluation found. High-ability students used it to verify work strategically, while low-ability students often used it to skip cognitive effort, potentially widening gaps.
Ministère de l'Éducation nationale — Senegal's Ministry of National Education will give 105,000 teachers AI training under the FORCE N programme, Minister Moustapha Guirassy said on 30 September 2026. The Université Numérique Cheikh Hamidou Kane runs it with the Mastercard Foundation, building artificial intelligence skills nationwide. Guirassy also handed 5,000 computers to secondary pupils.
Government of Punjab — Punjab's AI Curriculum and Training Initiative will deliver AI training for teachers to 160,000 educators. Ali Dar, adviser on AI to the Chief Minister, unveiled the classroom skills drive; no schedule or cost has been published.
University of Eswatini (UNESWA) — Generative AI training programmes at Eswatini's UNESWA AI Academy produced 480 graduates in 2026. 350 completed Mastering Generative AI for Professionals and roughly 130 finished Generative AI for Educators, out of over 570 sign-ups for these applied AI skills courses.
Banco Nacional de Guinea Ecuatorial — Banco Nacional de Guinea Ecuatorial has sent the entire payroll, every department included, through AI training for bank staff, delivered by BANGE Business School with Spanish education group CEF.-UDIMA. Jorge Lucio Sánchez Galán, who holds a doctorate in the subject, leads sessions covering banking applications, automated workflows, data work, fintech, cybersecurity, risk and responsible handling of digital tools. The lender expects this AI upskilling to surface efficiencies, lift output and quicken customer service across Equatorial Guinea.
University of Patras (KEDIVIM) — Greece's first university-certified cohort in AI in public administration wraps coursework this month and next, having started 29 June. The University of Patras's lifelong-learning centre backs the qualification, from the nation's first distance-based degree in this field. It sets a new benchmark for government AI credentialing nationwide.
OECD (PISA 2025) — The OECD's PISA 2025 results show 47% of Mexican 15-year-olds use AI chatbots for learning at least weekly, against 46% across the OECD. They exceed the OECD average for drafting schoolwork (35% against 29%), summarising texts (34% against 30%) and first research on a new topic (40% against 31%). Pupils avoiding AI for some tasks scored about 20 points higher in science, though the OECD says this does not prove AI weakens results.
The G8 industry bodies of Guernsey — Guernsey's G8 industry bodies drew over 1,000 sign-ups in six days for the island's first AI Sprint, a six-week AI skills training scheme running 1 October to 12 November. More than 50 free events, from talks to hackathons and clinics, welcome beginners and executives in regulated fields, offering artificial intelligence training island-wide.
OECD PISA 2025 — The OECD's PISA 2025 results show half of Dominican 15-year-olds (50%) use AI chatbots for learning at least weekly. Preliminary research on a new topic (44%), summarising assigned reading (40%) and drafting schoolwork (36%) lead, while only 9% never or almost never use chatbots for any school tasks PISA asked about.
OECD (PISA 2025) — The OECD's PISA 2025 results show 46% of students consult AI chatbots at least once a week. Such frequent student AI use was captured by the assessment's learning questions.
KPMG Canada — KPMG Canada polling from 22 July to 5 August 2026 found three-quarters of Canadian students use generative AI for schoolwork. Among educators, 59% doubt the system prepares students for AI-heavy jobs and 72% call such student AI use cheating, against 50% of students. Just 53% of students got formal responsible-AI teaching.
CARNET — CARNET, the Croatian Academic and Research Network, has finished a school AI curriculum spanning primary year five to the end of secondary, closing with 'Umjetna inteligencija: od koncepta do primjene' for third and fourth years. Uptake in 2025/2026 reached 121 primary and 84 secondary schools — extracurricular lower down, an elective above — at 35 hours a year. AI teaching in schools grew from CARNET's nearly 16 million euro project 'Podrska primjeni digitalnih tehnologija u obrazovanju - BrAIn'.
Bile Initiative / Tabaarak ICT Solutions — Tabaarak ICT Solutions and SomaliREN enrolled 160-plus final-year ICT students from 22 Mogadishu universities in round three of the Somali-run Bile Initiative, an AI training for university graduates scheme also covering cloud computing, software development and analysis. Hormuud Salaam Foundation funds it, steering graduates toward tech jobs. Women form 43 per cent of the cohort — an AI skills programme for ICT students too.
Mogadishu University — Somalia now has its first cybersecurity and AI degree, launched through Mogadishu University's Faculty of Computer Science and Information Technology. The new qualification targets Somali graduates entering emerging-tech careers, part of a push to grow local capacity for the country's expanding digital economy.
World Bank / Teachers in Sierra Leone — World Development Report 2026 counts more than 500 Sierra Leone teachers on TheTeacher.AI, a WhatsApp-based AI chatbot for teachers covering lesson plans, subject content and differentiated teaching. Blinded evaluators judged TheTeacher.AI answers more helpful and relevant than standard web search, while each reply needs 98% less data than a typical web page, so the teaching bot stays affordable on mobile.
University of Ghana — Ghana's University of Ghana will require AI training in higher education for every incoming undergraduate, folded into a compulsory Level 100 module from next semester. The One Million Coders Programme supplies four credentials: Prompting Essentials, AI Essentials, AI Professional and Digital Literacy. Ghana's Ministry of Communication donated 2,000 laptops to back the mandatory AI education drive.
Lebanese University — The Lebanese University has now run 430 professors and researchers through AI training for university faculty, 260 of them in a second round split across 13 groups. The courses cover applying artificial intelligence in teaching and research, and sit inside Supporting Integration of AI in Higher Education in the Middle East, a French Ministry for Europe and Foreign Affairs project reaching 13 universities across Lebanon, Egypt and Palestine. A third and closing round was set for early September.
Cambodia Academy of Digital Technology — 143 Club of Cambodian Journalists members, pulled from 82 outlets, sat an AI training for journalists course the Cambodia Academy of Digital Technology ran jointly with the club, titled Artificial Intelligence and the Future of Digital Journalism. Organisers pitched the newsroom AI skills session as a productivity boost for the national digital-news trade, per the Ministry of Post and Telecommunications.
Saudi Development and Reconstruction Program for Yemen — The Saudi Development and Reconstruction Program for Yemen has opened Digital Empowerment of the Yemeni Teacher, which extends AI training for teachers to 500 participants and 30 educational supervisors across Aden, Abyan, Hadramawt (valley and coast), Al-Mahra and Socotra. Two strands: artificial intelligence with specialised content generation, equipping teachers to apply AI to lessons, plus digital educational content design. These classroom AI skills are delivered with the Creators Centre for Studies and Training at King Abdulaziz University.
Maxis — Maxis extends AI skills training to over 32,000 pupils and teachers in 400 Malaysian schools via the enhanced 2026 Program GenAI Maxis eKelas, with the Education Ministry's Educational Resources and Technology Division. The programme's 2025 AI literacy pilot trained 22,000 participants, posting an 80 per cent median gain in AI understanding.
University of Papua New Guinea — A School of Business and Public Policy poll at University of Papua New Guinea, sampling 77 students, put daily AI use by university students at 42%, plus 30% logging in weekly — over seven in ten engaging at least once a week. ChatGPT led with 85% adoption, ahead of Microsoft Copilot (44%) and Google Gemini (23%). The unfinished research also found 74% wanting clearer campus rules, reflecting this student AI adoption trend.
Damascus University — AI use among medical students in Syria runs at 57% for study preparation and 40% for coursework on a 400-person sample of undergraduates and graduates. Clinical years outpaced graduates at revising and drafting. Optimism outran self-interest: 76% expected better residency training, 71.8% wanted institutional rules, yet 25.5% foresaw career gain, with women and researchers voicing ethical doubts. Fieldwork behind this student AI adoption ran April and May 2025 at public medical faculties.
Ceibal / ANEP — Three in four Uruguayan primary and secondary teachers now use AI in teaching day to day, finds a national Ceibal study published by education authority ANEP. Combining the National Teacher Survey and fieldwork, it found chatbots the most used tool, mainly for generating materials, designing activities and assessments, and preparing content.
Ministry of Education and Science of Tajikistan — Tajikistan's Ministry of Education and Science reports 468 students enrolled in AI higher education, as eight universities and institutes run a full artificial intelligence specialisation. Initial graduates, including the Technological University of Tajikistan's first four-year cohort, are due in 2026. University AI training spans machine learning, data analytics, algorithm design, AI system design, and image and speech recognition.
Eurostat — Generative AI use in Cyprus runs at 76.5 per cent among 16 to 24-year-olds for 2025, Eurostat figures show, comfortably ahead of the European Union benchmark. Across the EU the reading for that age band was 63.8 per cent over the same year. AI uptake among young people peaked in Greece at 83.5 per cent, then Estonia on 82.8, the Czech Republic 78.5.
In 2026, 95% of UK undergraduates report some AI use and 94% use generative AI for assessed work. Overall use was 66% in 2024.
Only 18% of US K-12 teachers had received any formal guidance on using AI, and under 1 in 10 had task-specific guidance (tutoring, grading, feedback), the 2026 Gallup/Walton 'Teaching for Tomorrow' survey (n=2,069, Feb-Mar 2026) found. Higher-needs schools lag wealthier ones, and adoption is outpacing policy.
The OECD's flagship Digital Education Outlook 2026 (January 2026) weighs AI's short-term task gains against what learners retain, and compiles member countries' data on school AI adoption. It is 2026's authoritative policy frame for education AI.
Cedefop finds AI job exposure lower on average in VET occupations than non-VET ones and projects a fairly modest 5% AI-driven employment decline in the EU. It still urges giving VET students AI skills, since graduates without them may look less employable.
Multiverse, the English apprenticeship provider training workers in AI and technology adoption, raised $70 million in May 2026 at a $2.1 billion valuation. Schroders Capital led the round, a $400 million rise on the previous valuation.
India's Cabinet approved the IndiaAI Mission in March 2024: ₹10,371.92 crore (about US$1.25 billion) over five years, with AI skilling among seven pillars. IndiaAI FutureSkills widens AI courses from undergraduate to PhD level and opens Data and AI Labs in Tier 2 and 3 cities.
Singapore's National AI Impact Programme targets AI skills training for 100,000 workers by 2029 and support for 10,000 enterprises over three years, including sector-specific courses for HR, legal, accounting, manufacturing and other non-tech professions. Budget 2026 set aside over S$400 million for an Enterprise Workforce Transformation Package, and Singaporeans on selected SkillsFuture AI courses get six months of free premium AI tools.
GTS Alive — GTS Alive, which issues ISIC cards, polled 530 sixth-formers in the Czech Republic in June 2026 and recorded 90 percent leaning on student AI use for coursework, against guidance at a fifth of schools. Roughly 80 percent of the year's leavers revised that way. Two thirds turn to ChatGPT and Google Gemini academically and almost half off duty; pupil AI usage goes largely unverified, with 44 percent verifying replies sometimes and 6 percent skipping that entirely.
Over 130 people in Rarotonga have now sat the Kia Orana AI Programme, workplace AI training run by a young Cook Islander and spanning government, financial services, hospitality and retail. The Infrastructure Cook Islands head office finished this AI upskilling in January 2026.
Université des Antilles — Université des Antilles now offers a new AI diploma for professionals on its Fouillole campus in Guadeloupe: the DU IIAM (Diplôme Universitaire in Initiation à l'Intelligence Artificielle Appliquée aux Métiers). Teachers, jobseekers and working adults with a Bac+2 can join the evening course, running March to July 2026. Graduates become AI Referents or digital transformation officers — a workplace AI training qualification.
Ministry of National Education of Senegal — Senegal's Ministry of National Education certified 3,467 public education agents of 3,909 candidates (88.69%) in Digital Skills (C2i-Edu) and Artificial Intelligence in Education (CIA-Edu). For AI certification in education, the AI Certificate in Education hit 93.1%. Planned: national rollout of these AI teaching credentials, adding beginner support and female-participation measures.
Education and Youth Development Bureau (DSEDJ) — Macao's Education and Youth Development Bureau (DSEDJ) will fund AI experimental classrooms in local schools from 2026/27, via a new grant under the Education Fund's School Development Subsidy Programme. Primary and secondary pupils will also get AI teaching materials and a smart teaching platform covering question banks, paper-setting and marking, all automated -- part of a wider AI in education push.
Maski (Bytefuse and Maskew Miller Learning) — More than 100,000 people are now using Maski, an AI education assistant that runs through WhatsApp, only months after Bytefuse released it in March 2025. Maskew Miller Learning helped build the tool, with R55 million in funding from Novus Holdings, and it now guides South African pupils across Grade 1 to Grade 12 through lessons matched to the national syllabus. No separate app is needed, since the AI-powered tutoring platform works directly inside the chat service.
Cetic.br — Generative AI in education has reached 70% adoption among Brazilian high schoolers online, per Cetic.br's TIC Educação survey, topped by ChatGPT and Gemini for homework research. Just 32% recall school guidance on safe, responsible use. The poll covered 7,476 students, 1,462 teachers and 945 managers across 1,023 schools nationwide.
National Centre for Curriculum (Saudi Arabia) — Six million-plus Saudi Arabia pupils follow a new AI curriculum in schools, live since 2025-2026. Four bodies built it: the National Centre for Curriculum, the Ministry of Education, the Ministry of Communications and Information Technology, and the Saudi Data and Artificial Intelligence Authority (SDAIA), packaging age-fit lessons into hands-on modules. It feeds Vision 2030's Human Capability Development Program, growing homegrown AI in education skills.
University of Sydney — Over 800 University of Sydney staff rely on Cogniti, a locally built AI assistant in higher education, crafting bespoke AI agents across fields like engineering and musicology. Australian Financial Review AI Awards judges gave it the Research and Education title. Such classroom AI assistant adoption remains rare campus-wide.
AI Academy Asia — AI Academy Asia's National AI Program will train 2,250 teachers and 250 AI specialists from 330 soums in 21 aimags over five years. UNICEF Mongolia and M Bank back the AI teacher training, a three-month course for ages 21 to 65. Training teachers in AI should narrow the digital divide and widen inclusive AI education.
AI chatbots help many U.S. teens with schoolwork. By late 2025, 54% of 13–17-year-olds said they had used one.
By 2030, 59% of the global workforce will need reskilling or upskilling, and 77% of surveyed employers plan to train staff to work alongside AI. Source: World Economic Forum's Future of Jobs Report 2025.
41% of VET policymakers, 36% of industry representatives and 21% of VET providers used or knew of AI in developing VET curricula, qualifications or occupational standards, OECD surveys of 25 countries (April–November 2025) found. Only 15% said relevant guidelines existed.
Across ten countries studied by the OECD, only about 20% of VET providers reported using AI in curriculum and qualification development. In the lowest-adopting country just 7 of 63 responding providers (13%) did.
Cedefop's pilot European Vocational Teacher Survey of 735 VET teachers in 23 EU Member States finds 29% using AI tools often or very often. Yet 31% say AI skills for teachers need major work in spotting relevant tools, and a similar share feel unready to judge AI's ethical risks and biases.
Beijing Municipal Education Commission — China's Beijing municipal government now leads the nation as the first provincial-level region requiring universal AI general education, with artificial intelligence classes launched across every one of the 1,400-plus primary and secondary schools, reaching 1.83 million pupils. Under a graduated approach, younger children get hands-on experience-based learning, middle schoolers pursue cognitive exploration, and senior students apply skills through comprehensive practice, with a floor of eight class hours set each academic year. Officials support delivery through 160 curriculum resource packages, training for 130 lead teachers, and a scheme linking urban schools with rural counterparts to widen access to AI courses in schools.
Ministry of Education (Børne- og Undervisningsministeriet) — More than 10,000 upper secondary students in Denmark now take part in AI in upper secondary teaching under a scheme run by the Ministry of Education, covering 75 schools and 442 classes across the 2025/2026 academic year. Four separate strands examine whether generative AI in classroom teaching changes how pupils write essays, how work gets graded, and how assessment itself functions in Danish, social science and natural sciences. Officials describe it as a deliberate experiment rather than a casual addition to lessons.
FCT (Fundação para a Ciência e a Tecnologia) / FCCN — Fundação para a Ciência e a Tecnologia has launched IAedu, a pilot for AI in higher education, giving Portugal's students, teachers and researchers free, centralised access to GPT-4o and Mistral via federated login. Built with Microsoft, it replaces individual subscriptions and aims to widen artificial intelligence in university research nationwide.
Only 0.3% to 5.5% of analysed training courses in Australia, Germany, Singapore and the United States delivered AI content, OECD analysis published in April 2025 shows, drawing on its 2024 training-supply study and 2023 policy questionnaire. Course supply is not yet keeping pace with demand for AI skills.
Singapore's SkillsFuture AI skills training nearly tripled. Participants in AI, cybersecurity and digital marketing courses rose from 34,000 in 2023 to 96,000 in 2024.
Advanced AI skills like machine learning and data science are sought after but scarce, held by about 1% of workers, the OECD reports. Meanwhile roughly a quarter of workers were already exposed to generative AI in 2022–2024.
Of 42,340 people who started a Wave 3 (2022–23) course in the UK's Skills Bootcamps programme, only 15,570 achieved a positive outcome, 37% of starters. While 54% of surveyed participants were promised a guaranteed employer interview, just 24% reported one offered or scheduled.
India's supreme audit institution found skilling scheme placements for just 2.318 million, or 41%, of the 5.614 million candidates certified under Pradhan Mantri Kaushal Vikas Yojana's short-term training. The ministry itself blamed the weak job placement on roles picked without skill-gap analysis or market-demand assessment.
Adult learning participation in the EU-27 favours the already-qualified: in 2022 just 25.1% of low-qualified adults took part, against 41.5% at medium level and 65.7% with tertiary qualifications. Adults holding a medium-level vocational qualification took part less than general-education peers at that level, 40.7% against 44.1%.
Three in ten (30%) U.S. public K–12 teachers report using AI tools at least weekly. They estimate saving 5.9 hours a week, about six weeks a school year.
Nám (Faroese Education Authority) and Mentamálaráðið (Ministry of Culture) — Four Faroese bodies — Nám, Mentamálaráðið, Undirvísingarstýrið and Fróðskaparsetur Føroya — co-authored a new AI strategy for the Faroese education sector. Targets for 2027, 2029 and 2032 anchor four themes: classroom practice, exams, skills-building, and research. The launch signals the islands' opening move on AI in schools policy.
Florida Board of Education — Florida's Board of Education has voted to fold AI policy for schools into every district and charter internet safety policy, governing student and teacher use, safety and data protection. Districts must publish their policies and an accessible list of approved AI instructional tools, then report those tools to the Florida Department of Education. Parents and guardians may opt a child in or out of direct AI interaction, with a non-AI alternative guaranteed; Florida College System institutions face parallel rules on academic integrity, privacy protection and AI literacy. This AI schools regulation takes force by July 1, 2027.
Ministry of Education and Culture (Mozambique) — The Ministry of Education and Culture plans to put Mozambique AI regulation for classrooms and campuses before the Council of Ministers by year end, Minister Samaria Tovela told the 4th National Conference on Distance Education in Maputo. Drafting is far along. The artificial intelligence rules will apply nationwide, from primary to higher education, with generative tools a chief concern.
Ministry of Education and Sport (Albania) — Albania's Ministry of Education has issued a guide permitting AI in schools across every pre-university level starting this academic year, naming ChatGPT, Gemini and Full Fact as approved tools. Staff can lean on them for drafting materials, setting and grading work, and language help across subjects from languages to maths, art and sport, while pupils are pointed toward research and creative projects rather than direct answers. Handing in AI output as one's own breaches the new academic-integrity rules on AI tools for classroom use, and a teacher must still check everything the software produces before it counts.
Los Angeles Unified School District — Los Angeles Unified School District, the second-largest US district, has suspended all student access to generative AI platforms for 2026-27, confirmed on 2 September, when its new Generative AI Ad Hoc Committee first met. Previously pupils 13-plus with digital citizenship training who had read the responsible-use policy could use approved AI tools.
New York City Public Schools — New York City public schools have paused student-facing generative AI for 2-K through eighth grade for 2026-27, a one-year moratorium the mayor's office announced. The policy covers nearly 600,000 students, about two-thirds of enrolment, and bans companion chatbots in every grade, while up to 50,000 high schoolers join five limited AI pilots.
University of the Virgin Islands — University of the Virgin Islands' Board of Trustees has unanimously adopted an AI governance policy for higher education, giving faculty, staff and students shared ethical-use guidance. A joint committee from all three groups wrote the rules on AI honesty in coursework, leaving departments a base for their own procedures.
UAE Cabinet, Ministry of Education (UAE) — The UAE Cabinet has approved an AI curriculum in schools reaching every public and private school, widening a rule that had covered only government schools. The subject enters the national framework instead of sitting outside lessons, running through all educational stages from the 2026-27 academic year. Alongside this mandatory AI school subject, 22,000 teachers and education professionals will learn to use AI tools for lesson planning, curriculum analysis, assessment and teaching.
Conselho Nacional de Educação — Brazil's National Education Council (CNE) has approved national AI guidelines for schools and universities that ban AI essay marking: no AI may correct, evaluate or suggest grades for essays or written exams at any level. Children in early-years education and primary school's first five years must not use AI tools without supervision, an AI detector's result cannot alone justify punishment, and automated grading cannot solely decide grades or progression. Bans apply on publication in the official gazette; institutions get 12 months for the rest.
District of Columbia Office of the State Superintendent of Education — The District of Columbia's Office of the State Superintendent of Education (OSSE) has issued a school staff AI policy for 2026–27, the AI Model Policy for Staff Use. Local education agencies may adapt the advisory text, OSSE's first staff AI guidance to them. A February 2026 OSSE survey found only 45% had a staff policy.
Florida Department of Education — The Florida Department of Education has proposed a rule requiring all 28 Florida College System institutions to adopt AI policies covering grading, course assignments and academic integrity for students, staff and campus guests. Education Commissioner Henry Mack approved it, and the State Board of Education considers it on 16 September. K-12 districts are also being told to add AI to internet safety guidelines, tell parents when an AI instructional tool is approved and let them opt children out, with school boards given until 1 July 2027; the State University System is drafting separate requirements for Florida's 12 public universities.
SIMAD University — SIMAD University has approved an Artificial Intelligence Policy Framework covering AI in higher education for teaching, research and administration. Drafted by the SIMAD AI Institute and endorsed by the University Council, the AI policy for universities aims to support responsible adoption while upholding academic integrity, transparency, privacy and oversight by humans.
Government of Seychelles — Cabinet has cleared Seychelles' National Artificial Intelligence (AI) in Education Framework 2026–2030, the country's first roadmap for AI in education in state schools. A four-tier National Teacher AI Competency Framework, a required AI micro-credential, and an AI Champion per school drive this push for AI adoption in classrooms, spanning teaching, leadership and administration.
Université d'Abomey-Calavi — Benin's Université d'Abomey-Calavi has adopted SIAUAC, a university AI policy pairing strategy, usage charter and ethics framework. The AI usage charter fixes principles, rules and duties for students, lecturers and administrators. The ethics framework adds obligations, technical requirements and governance wherever AI touches teaching, research, innovation or administration. By 2030 it wants standing as an African benchmark for artificial intelligence that is responsible, inclusive and sovereign.
Caribbean Examinations Council — The Caribbean Examinations Council is dropping school-based assessments in non-practical subjects for Paper 032, sat under invigilation, blaming fast-spreading generative artificial intelligence; the CXC AI policy came from Barbados, 16 July 2026. Losers include Mathematics, English, Caribbean History, Social Studies and Principles of Business, while Agricultural Science, Visual Arts, Music and other hands-on courses retain theirs. Regional exam AI rules hit CAPE from May-June 2027 and bind CSEC by May-June 2028. Registrar Dr Wayne Wesley called qualification integrity non-negotiable.
Norwegian Government (Kunnskapsdepartementet) — The Norwegian government said on 19 June 2026 AI in primary school should generally be avoided, and told the Directorate for Education and Training to issue national guidance before the school year. Year 1–7 pupils should not use AI for schoolwork themselves; lower-secondary classes may trial it cautiously once teachers are competent. All are advisory.
Ministry of Education and Technical Education (Egypt) — Egypt's Ministry of Education and Technical Education and UNESCO launched a national AI competency framework for teachers on 3 June 2026, localising UNESCO's global model. The teacher AI standards set the knowledge, skills and ethics for responsible classroom use; UNESCO counts the country among the first with a nationally contextualised version.
Universidad de la República (Udelar) — The governing council of Uruguay's public University of the Republic (Udelar) has approved university AI principles institution-wide. The document sees AI helping academic and administrative work and asks for transparent, ethical use; it puts 'general interest and centrality of the human person' first.
Ministry of Education of the Russian Federation — Russian schools add an Artificial Intelligence profile to advanced-level computer science lessons from 1 September 2026, per Education Minister Sergey Kravtsov. The AI school curriculum sits within in-depth computer science study. Textbooks for artificial intelligence teaching, covering grades 5-6, 7-8 and 9, joined the Ministry's federal list in February 2026.
Ministry of Education — Ethiopia's Ministry of Education is drafting a National Artificial Intelligence and Education Policy with UNESCO, covering AI in education policy nationwide. UNESCO's international experts are helping shape ethical, inclusive standards for AI adoption across education levels.
Government of Djibouti — The University of Djibouti gains a national Master's degree in Artificial Intelligence and Data Modelling, regulated by President Ismail Omar Guelleh's Decree 2026-092/PR/MENSUR. The two-year AI higher education programme, worth 120 ECTS credits, spans deep learning, machine learning, natural language processing and postgraduate AI study for health, banking and port logistics.
Ministry of Education (UAE) — The UAE Ministry of Education's guide on safe, responsible classroom use of generative AI in schools now has a first annex approving four platforms under set controls: ChatGPT, Microsoft Copilot, Google Gemini and Anthropic's Claude. Use of the approved platforms requires human review of outputs, fact-checking and academic integrity, the ministry says.
Higher Attestation Commission of Belarus — Belarus's Higher Attestation Commission has made AI disclosure in dissertations compulsory. Deputy chair Denis Kapskiy says the descriptive section must list the prompts an author ran and where artificial intelligence was involved. Antiplagiat, one of the standard packages used, flags the specific elements where neural networks were applied. Several 2025 dissertations already used machine learning, Kapskiy notes, with declared AI prompts and reasoned argument.
Ministry of National Education (Türkiye) — Türkiye's Ministry of National Education has launched YAZEK, an online system requiring teachers and ministry units to file an AI ethics declaration for school AI work. The companion Ethics Guide for AI Applications in Education sets when declarations apply, defines violations and reporting routes, and creates provincial and district ethics boards plus school AI-ethics teams.
England has created 'apprenticeship units', a new statutory class of short credential: 30 to 140 hours over 1 to 16 weeks, tied to existing occupational standards. Of the first ten units, published in April 2026, three cover AI leadership at Level 5, each 30 hours with up to £750 funding.
NCVET placed India's first AI skills qualifications on the National Skills Qualifications Framework in July 2026, recognising NIIT Foundation Awarding Body (Dual). The AI-tool pair are Fundamentals of Digital Productivity with AI Tools, NSQF Level 3, and Fundamentals of Data Management with AI Tools, NSQF Level 4.
The US Department of Labor's AI Literacy Framework, issued in February 2026 as Training and Employment Notice 07-25, targets state workforce and apprenticeship agencies, community and tribal colleges, Job Corps centres and state CTE directors. It defines AI literacy as grasping AI tools' capabilities and limits, applying them to daily tasks, giving clear instructions, checking output accuracy and following workplace rules.
Singapore combined SkillsFuture Singapore and Workforce Singapore into one statutory board, the Skills and Workforce Development Agency (SWDA), operating from 1 July 2026. The Ministry of Manpower and Ministry of Education oversee it jointly.
Cabinet of Seychelles / Ministry of Education — Seychelles' Cabinet has approved the National AI in Education Framework 2026–2030, requiring an AI micro-credential of every teacher. Schools each get an AI champion under a four-level teacher AI competency framework; a national steering committee will list approved and banned tools, and final grading and other high-stakes decisions keep human oversight in early phases.
Higher Education Authority — Ireland's HEA has issued a national policy framework guiding generative AI in higher education. Five pillars underpin it: integrity in academic work, equity plus inclusion, human oversight, data privacy and governance, and sustainable pedagogy. It steers responsible, values-led AI adoption in universities nationwide, rather than dictating one method.
Université de La Réunion — Université de La Réunion's Conseil d'Administration approved a Charter on the Use of Generative Artificial Intelligence, governing generative AI use in higher education across the institution for students, staff and faculty. Framed around responsibility, transparency and respect for academic values, it aims to support enriched learning rather than restrict it. The charter sits alongside internal rules and the wider national and European legal backdrop set by GDPR and the AI Act, supplementing but not replacing them, extending a broader generative AI policy at universities.
DRANE Guyane — DRANE Guyane, the region's academic digital directorate, has issued AI ground rules for schools in French Guiana, drawn from over 500 contributions gathered in a January to May 2025 consultation. Pupils gain unsupervised access to artificial intelligence tools only from Year 9 (4e) onward; younger students remain restricted. Data protection, transparency and environmental cost sit alongside pedagogy as core principles, and unauthorised generative AI use on homework counts as fraud.
Ministry of Technical Education, Vocational Training and Apprenticeship — Togo has begun shaping an AI in technical and vocational education strategy, steered by METFPA alongside the International Organisation of La Francophonie (OIF), launched in Lomé. A stakeholder panel is drafting the roadmap through consultation, weaving artificial intelligence into vocational training, with a completed version due early 2026.
Ministero dell'Istruzione e del Merito — A national blueprint for AI in schools now guides Italy's Ministry of Education and Merit on lawful, ethical classroom deployment, tutoring personalisation, admin task automation and data safeguards, tied to the Italian Strategy for AI 2024-2026. Individual institutions no longer improvise alone: the scheme binds education technology to coordinated national policy.
Barbados Accreditation Council — New guidance from the Barbados Accreditation Council (BAC), the Guiding Principles for the Use of Emerging Technologies in Post-Secondary/Tertiary Education, sets rules for AI in post-secondary/tertiary education. Explainability and human intervention must underpin automated calls, with assessment resting mainly on human judgement. Fairness, data security and equitable access complete this early marker of AI adoption in tertiary education.
UIN Syarif Hidayatullah Jakarta — Rector's Decree No. 127/2025 makes UIN Syarif Hidayatullah Jakarta the first Islamic university in Indonesia with formal rules on generative AI in education, spanning Generative AI and Large Language Models in academic and non-academic activities. Exams stay off-limits unless a faculty sets its own protocol; written work needs clear disclosure of AI contribution. Curriculum design, material prep and reporting remain open to generative AI use in universities.
England has a Level 4 'AI and automation practitioner' apprenticeship standard (ST1512), approved for delivery on 10 December 2025, for non-specialist operational staff using AI and automation tools rather than AI engineers. It typically lasts 18 months, with a 420-hour minimum, up to £18,000 funding per apprentice and Ofqual providing external quality assurance.
Ministry of Education Brunei Darussalam — A handbook titled 'GenAI Guidance for Education – Handbook Edition', issued by Brunei's Ministry of Education, sets AI in education standards for students and teachers, spanning safety, honesty and privacy. Its rules echo the AITI AI ethics guidelines and Personal Data Protection Order, embedding generative AI adoption within existing national policy.
Ministry of Education, Science and Technology — Tanzania's schools and universities must build their own AI in education policy, per National Guidelines for Artificial Intelligence in Education from the Ministry of Education, Science and Technology. Nested in the National Digital Education Strategy 2024/25-2029/30, the framework covers privacy, security, ethics and learner well-being, balancing responsible artificial intelligence policy with classroom tradition.
Université Virtuelle de Tunis (UVT) — A framework for ethical AI use in Tunisian higher education now has a second edition. The Virtual University of Tunis released version 2 of the Charter for Good Use of Artificial Intelligence at the Tunisian University on 19 December 2025, setting expectations for how staff and students handle machine-assisted work across teaching and scholarship nationwide. Human judgement, the document insists, must still outrank automated output whenever decisions get made. Its scope stretches from safeguarding personal data to guarding against dishonest submissions and demanding openness about when algorithms enter the process — the country's clearest statement yet on AI governance in academic research.
Colruyt Group (Boir) — Colruyt Group's drinks webshop Boir introduced Babette, an AI sommelier choosing from nearly 2,000 products by budget, occasion or taste. Customer service and sommeliers take tricky questions.
Mercadona — Mercadona has over 500 engineers on Atenea, an in-house platform of agentic AI for software development. Specialised AI coding agents draw on Mercadona's quality standards, tools, processes and knowledge.
Lawson — Convenience chain Lawson has debuted its first large-scale AI-conceived sweets and bakery items, built from flavour pairings a generative AI system dreamed up before staff refined them into sellable products. The lineup spans a pickle-lemon tart, a sake-fermented red-bean-and-yoghurt bread, and a parfait-style salad in Japanese presentation. About 4,700 shops across Kanto and Koshinetsu began selling all three, with the AI-generated flavour concepts serving as the starting point for each recipe.
FamilyMart — FamilyMart's debut in AI product development in retail: a caramel-sauced sweet potato canele, 285 yen, now stocked nationwide. An algorithm processed two years of sales and customer data into about 1,000 candidate recipes; humans chose this one. 10 AI-generated confectionery items follow by February 2027.
IKI — IKI is extending AI promotion planning to 250 supermarkets in Lithuania with SymphonyAI's Promotional Evaluation and Promotional Planning products. Category trials lifted promotional forecast accuracy by more than 10 percentage points and cut empty shelves, vendor-reported figures. Category management head Klavs Berzins says buyers now foresee demand, the point of retail demand forecasting.
Macy's — Macy's is scaling AI inventory replenishment from pilot to wider rollout, overlaying AI forecasts on replenishment to raise in-stock rates and inventory efficiency, COO and CFO Tom Edwards told the 10 September earnings call. The forecasting layer sits inside a supply-chain-led transformation plan expected to save $235 million by 2026.
Msheireb Properties — Msheireb Properties has started phase one of Rashid, an AI digital concierge answering from 15 interactive smart kiosks in Msheireb Downtown Doha. Rashid gives wayfinding plus live detail on events, parking and nearby restaurants, day and night, taking spoken Arabic and English. Hewlett Packard Enterprise built the smart city AI assistant on Fanar, Qatar's Arabic large language model, with MEEZA running the platform.
Carrefour Belgium — Two Waterloo sites, a hypermarket and a franchised Carrefour Market, mark Carrefour Belgium's rollout of AI shopping trolleys. Smart carts for grocery shopping, built from Instacart's Caper Cart, detect products on contact and weigh loose produce automatically.
AEON360 — AEON360, AEON Malaysia's data and AI unit, uses AI customer profiling to sort shoppers by life stage, not transactions alone, pooling retail, mall and financial service records. Customer profiling with AI feeds a recommendation engine weighing transaction data, buying frequency, engagement and preferences, so personalised suggestions evolve with shoppers.
Allegro — Allegro has made a voice AI agent, built with ElevenLabs, the first point of contact on its Allegro Delivery hotline after a testing period. The agent answers questions on parcel collection and delivery status, and hands more complex cases to a hotline employee.
Seznam.cz — Seznam has switched on AI agent shopping across Zboží.cz, its Czech price-comparison platform, as a pilot. A signed-in shopper's agent completes the purchase, cutting checkout to two clicks and skipping manual delivery entry. Compatible retailers are joining gradually, and Seznam plans to extend this automated buying feature sitewide.
Högskolan på Åland — Högskolan på Åland's Light in the Dark project has produced Tourism Business Idea Generator, an AI for tourism businesses tool with a €20,000 budget. Small and medium operators around the Baltic Sea can use it for custom low-season products and marketing content, no prompt engineering required. The AI tourism idea tool targets quiet-season gaps with tailored suggestions over generic templates.
Best Buy — Best Buy launched 'Ask Blue', a conversational AI assistant drawing on product details, reviews, availability and pricing, helping shoppers compare items, check compatibility and answer questions, and routing them to self-service or live support. Incoming chief executive Jason Bonfig announced the phased rollout, without technology or adoption figures.
Sainsbury's — Sainsbury's says it will extend Facewatch facial-recognition technology for loss prevention to about 200 stores in 2026, following a false shoplifting accusation against a customer. The retailer put the incident down to human error, not the technology, and says the system is 99.98% accurate, with a trained manager reviewing every match.
Dollar General — Dollar General has deployed RELEX Solutions' AI platform for replenishment, allocation and demand forecasting across 21,000-plus stores and 34 distribution centres, uniting forecasting and replenishment in one system. Jeff Vaughan, the retailer's SVP of global inventory management, said it gives teams more visibility across the network.
Marks & Spencer — Marks & Spencer has put CattleEye's AI welfare monitoring on all 46 farms in its dairy milk pool, with cameras and video analytics scoring each cow's feeding, lying time, mobility and body condition into a 'cow comfort score'. M&S claims a retail first on dairy farms; earlier trials prompted new bedding, feeding schedules and cubicle designs.
The Home Depot — Home Depot's Magic Apron AI assistant is now live in all 2,000-plus US stores with localised store knowledge, helping customers and associates find products, get answers on specific items and plan projects in the aisles. Shoppers reach it via the Home Depot app or an in-store QR code and can type, dictate, upload photos or use their preferred language. The retailer says it already fields millions of questions a month.
Lotte Department Store — Lotte Department Store's in-house 'Brand AI' analyses more than 4,000 brands for AI merchandising strategy and emerging-brand scouting, the company says. Launched in late July 2026, it processes the data buyers need into visuals, with a natural-language chatbot due in the second half of 2026.
Carrefour Belgium — Carrefour Belgium has launched an AI shopping assistant inside ChatGPT, built with Mealz.ai, that places grocery orders, checks stock and draws up shopping lists. It proposes priced meal plans, remembers dietary preferences and connects to the Carrefour Bonus Card, with home delivery or in-store pickup, following Carrefour France's earlier launch.
JAZA Supermarkets — Discount chain JAZA runs AI for pricing in retail alongside stock control and delivery routes across Kenya, a 2026 KPMG case study found. Real-time rival comparisons and behaviour tracking feed a central warehouse linked to branches for faster restocking. Founder Willy Kimani said the same data guides marketing and new site choices, echoing AI-driven inventory management.
Canadian Tire — Canadian Tire is giving corporate staff Microsoft 365 Copilot alongside AI pricing and retail-intelligence tools. Its DaiVID platform sets prices and promotions, while MOSaiC, built with Microsoft on Azure, had found over 1,000 customer 'life occasions' and 180,000 demand signals by the first quarter of 2026. Staff numbers were not disclosed.
Rotana — Rotana has introduced H2O Hospitality's AI hotel check-in at 15 Rotana-branded Abu Dhabi properties, 4,240 rooms across Rotana Hotels & Resorts, Arjaan and Centro. Guests send passport details and a face photo over WhatsApp a day or two ahead, and facial recognition confirms them at the front desk. Both firms are discussing expansion outside Abu Dhabi.
AllSaints — AllSaints has taken a Klaviyo-based customer engagement platform live worldwide, using AI to tailor content and open new customer channels. The retailer calls the platform the first project of its new engineering team at its Global Capability Centre in India. No store, country or customer figures were given.
Just Eat Takeaway.com — Just Eat Takeaway.com has taken its AI voice assistant Europe-wide after UK and German launches, letting customers search restaurants, grocers and shops by speech on iOS and Android. Early UK company data show users finding things about 40% faster than with text-only AI and 30% likelier to choose a recommended dish or restaurant; no independent evaluation exists.
Kroger — Kroger now runs 500 Simbe 'Tally' shelf-scanning robots across its stores, the 500th going live at its Smith's Marketplace banner, Simbe says. The autonomous robots patrol aisles checking on-shelf availability, price accuracy and inventory; Kroger has published no figures for its fleet.
Morrisons — Morrisons has begun putting Everseen's 'Evercheck' computer-vision system on self-checkouts in up to 200 stores, where real-time image recognition spots possibly mis-scanned items and prompts customers to correct them. The retailer says it checks transactions rather than people, with no facial recognition, and aims to keep queues moving and cut staff interventions.
Asda — Asda has extended Auror's AI retail-crime intelligence platform to every store after a 10-week pilot in 30 London stores. Asda's own pilot figures, not independently evaluated, show over £100,000 of goods recovered, incident reporting cut from about 25 minutes to under five and crime-prevention rates up 55%. Chief operating officer David Lepley said staff now report incidents and identify repeat offenders faster.
B.TECH — AI pricing electronics retail took hold at B.TECH, Egypt's biggest electronics chain, after a two-year McKinsey overhaul of pricing, logistics and credit. Conversion rose 14 per cent on a new AI-driven e-commerce site, while an AI logistics hub sped deliveries. Mylo now scores borrowers, chases collections and flags fraud via AI-driven e-commerce and credit scoring methods, reaching hundreds of merchants across over 15 Egyptian cities in year one.
Rohlik Group — Rohlik bills the new offering as among Europe's earliest AI shopping apps living inside ChatGPT. Ordinary conversation covers finding goods, turning up recipes, assembling or revising a basket, and steering a whole order to completion. Rohlik's in-house Model Context Protocol server enables the hookup and reaches others, alongside Maia, the grocer's own conversational shopping tool; no user or order counts are given.
Swisscom — Swisscom is piloting an AI system named KIRA that analyses customer conversations in three Swiss retail branches, located in Bern, Aarau and Lausanne, as part of a broader push toward AI in retail customer service. Motion sensors monitor shopper movement while microphones capture sales discussions with customer consent, and anonymised transcripts are then reviewed through AI-driven retail analytics to surface what shoppers are actually seeking. The trial is scheduled to last two to three months, and if it proves effective, Swisscom intends to roll KIRA out to 10 to 15 branches by the end of the year.
Pick n Pay — Pick n Pay's asap! arm has switched on Penny, an AI grocery shopping companion built on Google's Gemini AI models, billed as a South African first. Customers speak, type or send a photo instead of clicking menus; the AI-powered shopping assistant draws on Smart Shopper loyalty history to shape suggestions.
Tourism Development Department — The Tourism Development Department and Dynamik Technologies have switched on Nakhoda, a booking hub whose AI trip planner drafts a tailored route around a holidaymaker's likes and wallet. Operators use the same website to post services, build packages and collect money without cash. Nakhoda arrives before Visit Brunei Year 2027, handing tourists AI itinerary planning early.
Mironet — Mironet, a Czech retailer, now leans on five home-grown AI agents in e-commerce — DÉDAL, ATLAS, IRIS, FÉNIX and ARGOS — covering code, analytics, mail, freight and security. Work once eating a whole shift of hand-built exports and spreadsheet formulas now wraps up daily. Mironet reckons the agentic automation platform worth up to 19 million crowns yearly, equal to 20 percent of wage costs and to 24 posts, contractors included.
Kenzz — Pricing guidance, demand forecasting and market intelligence reach sellers through Bosla, the AI seller platform built by Kenzz for Egypt's first mass-market e-commerce site. Shoppers interact with Morgan El Kenzz, an in-app assistant suggesting products and fielding support questions, escalating harder cases to staff. Kenzz's figures show conversions climbing 4.5 to 5.3 times above ordinary browsing, alongside a near-25 per cent gain in first-contact resolution, within the company's AI-driven e-commerce ecosystem.
Anguilla Tourist Board — The Anguilla Tourist Board has put an AI concierge for tourism into visitors' hands with the 27 May 2026 launch of its Xcape destination app, part of its "Taste, Feel, Live" summer campaign. The concierge weighs weather, safety concerns and user ratings, flagging unsuitable sea conditions or tides and suggesting safer options, alongside GPS routes, nearby beaches and sights, and where to eat. Wilson Jno Baptiste and Gavin Jean built the platform; Destination Experience Manager Keri Rogers said the idea came from wanting to help visitors find their way, and anyone can download it at no cost from the App Store and Google Play.
Rohlik Group — Rohlik Group has wired card payments straight into Maia, making AI grocery shopping one unbroken flow: pick items, check out, approve the charge on a saved card, never quitting the thread. Maia also takes spoken orders via Siri and Google Assistant, plus product hunts, recipes, lists and complaints. An MCP server lets ChatGPT and Claude plug in behind secured OAuth verification; the voice commerce assistant operates wherever Rohlik claims a Net Promoter Score beyond 90.
CEMACO — Guatemalan hardware, home and decoration chain CEMACO has launched MACO, an AI shopping assistant that helps customers choose purchases. Shoppers reach the virtual adviser through in-store digital kiosks, WhatsApp or the chain's website.
Shoprite Holdings — ShopriteX built Pixie, an AI shopping assistant for the Shoprite Group's Sixty60 app. Machine learning lets it study buying habits and flag restocks, unfamiliar products and deals automatically. Now in beta for Xtra Savings Plus members, it makes Sixty60 the country's first online grocer with a personalised AI shopping tool.
METROPOLE optical salon chain — METROPOLE is Latvia's first optical salon chain offering AI eye screening and digital vision testing, now across 18 salons. The chain's AI-based eye disease detection rates risk of three common conditions early, while digital testing enables remote optometrist consultations. Upgrades have cost METROPOLE over €2 million since 2024.
Trial Company — Every one of Trial Company's 264 super centre branches has gone live with CIX-自動発注 (CIX-Auto Ordering), finishing a nationwide push for this AI ordering optimisation system. Machine learning paired with digital twin simulation drives down surplus stock by close to 30%, while liberating 180 person-hours of labour monthly at each branch. Automated inventory ordering has become the chain's standard practice.
Shimamura — Shimamura's group EC sites and store app now run on a Neural Group engine built for AI product recommendations in fashion retail. Tens of thousands of product images get cross-matched with live stock, buyer history and web trends to surface tailored picks, while the app flags nearby shelf availability instantly through AI-driven personalised shopping recommendations.
Loblaw Companies — Loblaw has put a PC Express app inside ChatGPT, letting Canadians order groceries and plan meals with their local store by chat. OpenAI models already run behind its own tools, including Robin, an AI assistant for store owners and managers, and agentic supply-chain systems for inventory accuracy and logistics. Corporate staff will get ChatGPT Enterprise.
Metro Slovakia — Metro Slovakia runs an autonomous store retail site in Mlynky, part of the Môj obchod partner network, staffed by day and switched to unattended AI-run operation once employees leave. Computer-vision cameras log every item taken from shelves, giving shoppers 24/7 access without a cashier on duty. Staff return each morning, so the shop is only self-operating outside normal opening hours.
Pyramid Continental Hotel — A humanoid figure named Ayen now greets diners at Pyramid Continental Hotel in Juba, marking South Sudan's first AI robot waitress hospitality installation. She wanders the dining floor, converses with patrons, and reinforces food and beverage staff, sharpening the overall guest experience — billed as the nation's debut humanoid service robot hotel case.
Allegro — Allegro has opened its AI selling assistant to every marketplace partner after positive early feedback, and made its shopper-facing AI Assistant fully available in the Allegro app after well-received tests. The Polish marketplace, with over 80 million products from more than 150,000 sellers, announced both with its first-quarter 2026 results. It runs close to 100 AI projects and aims for AI to underpin about 40% of its technology portfolio by year-end.
Woolworths Group — Woolworths has made its Olive tool an agentic AI assistant for all roughly 200,000 group staff, live a week before an executive described it at a conference. Combining search and chat, it adds requested items to carts, suggests recipes from meal photos and offers cheaper or organic swaps, while in-house 'agentic judges' vet answers for accuracy and compliance first.
Aldar — Aldar has partnered with Visa to bring AI-powered property payments to the UAE, using an agentic checkout system called Visa Intelligent Commerce in its first live regional rollout. Speaking a request into an AI agent on the Live Aldar app or website triggers the transaction, with the customer's approval required before anything is charged for real estate service fees. An Emirates NBD Darna Visa Credit Card processed the debut payment.
Ditano — Thousands of fragrance and beauty products now get algorithm-driven suggestions on Ditano's site, where the Italian retailer has rolled out conversational AI in e-commerce built by TextYess, spanning both its online store and WhatsApp. Acting as an AI beauty consultant, the system walks shoppers toward tailored picks. Recovering carts customers abandoned, it converts at 11% and returns 128x ROAS, per figures the companies published.
Empik — Empik's AI gift advisor, built by Polish developer Zowie, lets shoppers describe what they want in plain speech or text instead of using category filters. The conversational shopping agent, live on the retailer's website, reads context and recipient details to recommend presents, replacing product search with natural conversation.
Olsson Parts — Olsson Parts now offers AI customer service through CAI (Customer Artificial Intelligence), a six-language chatbot on its site. Human staff still handle tricky queries; CAI covers the routine ones — the retailer's first AI-driven customer support tool.
Siam Piwat Group — Siam Piwat Group now runs AI in retail across its ONESIAM SuperApp, spanning Siam Paragon and ICONSIAM, on Google Cloud infrastructure. Vertex AI drives semantic search for shoppers, swapping older rule-based lookup for personalised recommendations, while a predictive segmentation tool targets marketing campaigns as part of the retail AI stack. Snap & Collect applies Vision AI to receipts, automating loyalty rewards and e-stamping without staff handling.
C&A — C&A holds a national first: Brazil's inaugural AI personal shopper for fashion, built into its website and app. Shoppers converse with the tool, receiving outfit pairings matched to taste and purchase intent. A R$300m digital budget for 2025 backs this AI-powered styling assistant, alongside a hyper-personalised homepage and design tools for C&A's creative staff.
Tupi S.A. — AI in retail operations now covers marketing, HR and collections at Tupi, a Paraguayan e-commerce firm. Google Performance Max lifted results 50 per cent, hitting 300 per cent ROI on one push; hiring software sifts psychological tests and forms into a shortlist. A collections chatbot settles 80 per cent of queries and runs payments unattended, marking an AI-driven retail business shift.
Kuningan City Mall — Kuningan City Mall now runs Airy, short for AI Robotic Kuningan City, an AI humanoid robot customer service unit at its concierge desk. Built with Indonesian vendors, it directs shoppers to tenants, deals and events, speaking English, Chinese and French. The move backs the mall's push toward a hi-tech venue, an early humanoid robot retail deployment in Indonesia.
TOPS (Central Food Retail Group, Central Retail) — Central Retail's food chain TOPS deployed AI stock checking retail via a fresh chatbot letting buyers confirm shelf availability nearby before a visit. Built on Google Cloud's Conversational Agents platform, it adds 24/7 chatbot support and natural Thai comprehension for steering shoppers inside branches.
Premium Nexus JSC — Over 500 CU stores and two distribution centres in Mongolia gain AI demand forecasting in retail via Premium Nexus JSC's deal with RELEX Solutions. One unified system now runs warehouse and shop-floor planning, producing automatic forecasts, tighter stock strategies and location-level KPIs through the AI-powered forecasting and replenishment system.
SCENTMATIC — Tsim Sha Tsui's new @cosme HONG KONG flagship, the brand's first shop outside Japan, now runs SCENTMATIC's AI fragrance system KAORIUM. It performs AI scent matching, turning fragrance into a 16-word Scent Index for description-based browsing. Choices sync to the @cosme app across online and in-store visits.
Sansiri Public Company Limited — Amazon Web Services underpins Sansiri's ChatX Assistant, Thailand's first large-scale generative AI in real estate, built on Anthropic's Claude via Amazon Bedrock. Buyers and staff get instant answers on financing, zoning and construction across 500-plus projects via this AI-powered home buying assistant. A linked OCR tool reads over 50,000 invoices monthly above 90% accuracy, lifting response speed 30% and satisfaction 25%.
bol.com — Inside bol.com's app, Cadeauhulp now handles AI gift recommendation by asking who the gift is for, the relationship, and budget, then pulling matches from the Dutch retailer's catalogue. Agency partner DEPT co-built the launch, bol.com's debut into AI-driven gift suggestions for browsing products.
Coppel — Coppel rewrites 123,000 product listings with FeedGen, a Google Vertex AI e-commerce tool, cutting an eight-day task to under five hours. The Mexican retailer's holiday-season revenue rose 27% and clicks 48%, gains attributed to automated catalogue optimisation. Figures are Coppel's own, unaudited.
Grupo Boticário — Conversion rose 46% within a month of Grupo Boticário launching its AI shopping assistant on the O Boticário app, built on Amazon Bedrock Agents and Anthropic Claude Sonnet 4. Buyers using the retail chatbot paid 7.4% more per basket than those bypassing it.
Majid Al Futtaim — Majid Al Futtaim is trialling Agent Pay, Mastercard's AI-powered payment tool, in Dubai with fintech firm Dataiera, the service's debut purchase beyond the US. Via this AI shopping assistant, customers let software hunt bargains, weigh alternatives and settle the bill. Agent-assisted checkout begins with VOX Cinemas ticket bookings.
Debenhams Group — Peak's agentic AI for pricing and stock now runs across Debenhams Group's brands — PrettyLittleThing, Boohoo, BoohooMAN, Karen Millen and Debenhams. Live data feeds autonomous decisions, moving inventory and resetting prices without human sign-off. This automated retail pricing tool eases merchandising pressure before the festive rush.
BUA Foods — BUA Foods, the Nigerian FMCG arm of BUA Group, now drives AI-driven sales automation using BeatRoute's Sales Force Automation Software and Customer App. Reps get smart scheduling, order optimisation and plain-language lookups against live figures, swapping out paper-based habits across the distribution footprint. This BeatRoute rollout, an AI-powered sales force automation tool, seeks tighter delivery planning group-wide.
Fleurop-Interflora (Schweiz) AG — Fleurop-Interflora (Schweiz) AG nearly doubled conversions after adding Flora, an AI shopping assistant for flowers on the Marple Companion platform. Shoppers pick arrangements via sliders and colour swatches instead of chat, buying 99% more often and spending 3% extra per order. The firm also piloted a bouquet subscription sold only through this AI-driven retail assistant.
President Chain Store (7-ELEVEN) — 7-ELEVEN has opened ninth-generation smart store X-STORE 9 at Taiwan's National Central University, using AI image recognition, skeleton tracking and ceiling LiDAR to identify over 6,000 products for AI grab-and-go shopping. Entry uses a transport card, and payment is at a self-checkout kiosk. The cashierless system, built with the Industrial Technology Research Institute (ITRI), serves 20 shoppers simultaneously on standard fixtures, not smart shelves.
TUI Nederland — TUI Nederland deploys Lisa, an AI travel agent for holiday bookings, as a lobby avatar handling opening customer chat. Destination, dates and party size get logged automatically, then routed to the branch so advisors focus on tailored advice. Trojan Monkeys, an AI content agency, built the retail assistant.
Tokopedia — Counterfeit product detection for e-commerce got a 40% drop in illegal listing reports after Tokopedia ran SmartShield AI through a six-month trial, VP of Trust and Safety Dinda Maharani said. Image recognition paired with text analysis lets the system flag fraud patterns across thousands of daily listings and pull fake goods automatically. Indonesia's marketplace now leans on automated fraud-spotting for online retail to keep its platform clean.
S-ryhmä (S Group) — S-ryhmä, Finland's largest retail cooperative, has introduced AI multilingual customer service for S-Etukortti loyalty-bonus guidance, covering 17 languages through a new web tool. An animated guide, Kielitietäjä-Eero, narrates explainer clips and switches tongues mid-sentence, powered by generative video technology. Helsinki's HOK-Elanto service points piloted this multilingual bonus rollout first.
Magazine Luiza — Brazil's Magalu (Magazine Luiza) offers AI commerce via 'WhatsApp da Lu', fronted by virtual influencer Lu. Customers browse, select and pay inside the app, guided by a conversational shopping assistant on a multi-agent AI system. Roughly one million regular buyers have early access.
Goddiva — Goddiva shoppers now chat with an AI shopping assistant, Goddiva AI, built into the British label's site and order platform. Powered by OpenAI GPT-5 with LangChain, it handles discovery, order status, returns and discounts, standing in for an in-store stylist across online fashion retail.
Natural Lawson (Lawson) — Six Natural Lawson outlets in Japan now offer an AI sommelier, KAORIUM for Sake & Wine, to guide customers toward suitable bottles. Shoppers enter their mood and flavour preferences via a tablet, and the system pairs a recommended wine with snacks available in the shop, using interface options in English and Chinese besides Japanese. Whether the AI-driven wine recommendation tool spreads to additional branches depends on how this trial performs, per Lawson.
Coles — Coles has become the first major Australian retailer to roll out ChatGPT Enterprise deployment at scale, moving toward full company-wide use by February 2026. Powered by OpenAI's GPT-5 model, the aim is faster research, lighter paperwork and sharper data insights — an early instance of enterprise generative AI in retail.
Snoonu — A new build of Snoonu's super app introduces Genie, an AI shopping assistant able to parse everyday requests, say a request for a skincare routine or a present idea, and surface matching products. Multiple coordinated agents power the suggestions for every user on the platform, extending Qatar's move toward AI-driven retail assistance.
Hyundai Department Store — Hyundai Department Store's generative AI shopping assistant HEYDI (Hyundai Personalized Lifestyle Design AI) is now live across every branch, outlet, website and mobile app in South Korea, following a phased rollout that began with foreign shoppers in July 2025 before reaching domestic customers three months later. The tool interprets natural conversation to suggest stores, restaurants, pop-ups and exhibitions matched to individual tastes, and can handle layered queries like gift ideas or planning a shopping route. The nationwide version merges both customer groups into a single integrated service, marking a shift from its earlier limited release toward full retail AI personalisation.
Maxim's Group — Maxim's Group has turned to AI customer data unification to merge fragmented records across its Hong Kong restaurant portfolio, which spans quick-service outlets to upscale dining rooms. Staff without technical backgrounds now query an agent built on Microsoft Copilot Studio using everyday language to pull tailored reports, while a second agent trawls thousands of files for marketing campaign details. Both feed into single customer profiles covering every visit occasion, an early instance of AI-driven unification of customer data reshaping how the group reads its dining network.
Hyphen Tech — Hyphen Tech's Cosebu Ecosystem opens an AI marketplace Burundi traders can use, debuting at the firm's Bujumbura hub. Merchants gain Cosebu Management AI, covering inventory, sales tracking and financial reporting, plus Cosebu Marketplace AI, powering discovery through AI recommendations. Kirundi, French and English speakers are served by this AI-driven commerce platform, built to draw Burundian youth into digital trade.
Chalhoub Group / FACES — FACES and Chalhoub Group built Layla AI, an AI beauty coach chatting bilingually in Arabic and English, pairing shoppers with fitting skincare, makeup, fragrance and haircare picks. This AI-powered beauty advisor drives 2.5-fold higher conversion, plus 7.4 extra browsing minutes. Chalhoub Group supplied these unverified figures.
Channel Islands Coop — A till-side scan now decides whether Jersey shoppers look old enough for restricted goods, after Channel Islands Coop fitted ten stores with facial age estimation technology at self-checkout points. The retailer calls it the island's first such trial; no photos or personal records are kept by the age verification system. Staff still oversee sales, but the software confirms eligibility first.
Core Innovations (Motor Oil Group) — Greece's first autonomous AI convenience store has opened in Nea Smyrni: Smartshop, run by Core Innovations, a Motor Oil Group subsidiary. This unmanned retail format uses computer vision and sensors to stay open unstaffed 23:00–07:00, with card payment and three-step checkout.
Silvie's Industrial Solutions — Ten seconds is now the wait for an industrial parts match at Silvie's Industrial Solutions, thanks to its new AI-powered product selection agent, Silvie AI. Customers scan a code and the multilingual system replies 24/7, while managers gain visibility into buying trends that let seasoned staff shift away from routine searches toward higher-value client work. The launch signals wider adoption of AI-driven product search across Guyana's industrial retail sector.
Multiplaza Bogotá — Colombia's first mall-based AI personal shopper has launched at Multiplaza Bogotá: MIA (Multiplaza Inteligencia Artificial), built with AI Findr. Shoppers get instant answers and tailored product picks through natural conversation — a retail AI shopping assistant setting a new benchmark for malls nationwide.
Colruyt Group — Halle shoppers can trial the AI-powered self-scanning shopping trolley, Smart Cart, now live at Colruyt Group's Lowest Prices branch, built in-house with EXPRESSO. Sensors and onboard processing let customers ring up goods and pay as they browse, an AI self-checkout cart retail approach that skips the till queue. A linked Smart Scale weighs loose produce with over 98% accuracy, and Colruyt calls itself the first Belgian chain to offer such a trolley.
Juewei Food — Juewei Food's AI marketing agent for retail: 3.1x sales, 1.8x click-through, 2.4x payment conversion, 47% fewer WeCom friend deletions — tested on millions of shoppers versus human teams. China's largest braised snack chain built this full-chain AI marketing automation with Tencent Smart Retail on Hunyuan and DeepSeek dual-model architecture, spanning strategy through execution. Juewei's own trial figures, unverified externally.
Grandiose — Grandiose, a UAE grocery chain with nearly 50 stores, has embedded GrandChef, an AI cooking assistant, inside its mobile app, converting recipes, dietary needs and meal ideas straight into shopping carts. Powered by Microsoft technology, this AI-powered chef stays limited to a pilot on the newest Android and iOS releases.
Lawson — KDDI and Lawson are trialling AI in convenience stores at 221 Lawson outlets in the Philippines, applying AI-driven store management to sales and inventory data for shelf layouts and order volumes. A separate system reads security camera footage for stock levels and checkout congestion. The venture targets efficiency and revenue gains.
MediaMarktSaturn — Round-the-clock answers on roughly 1,300 own-brand items now come from Smart Manual, MediaMarktSaturn's new AI chatbot customer service tool powered by Google's Gemini model, covering PEAQ, KOENIC and ok. product lines through manuals, technical data sheets and FAQ content. Debuting in Germany during August 2025, this AI customer support tool has launched solely on own-brand sites for now, with wider integration into the retailer's core webstores and mobile apps still to come.
Sheng Siong — All 83 Sheng Siong outlets in Singapore now run facial recognition CCTV retail surveillance against shoplifting, built with the Singapore Police Force under the STAR programme. Flagged repeat offenders trigger an instant phone alert to staff on re-entry. Sheng Siong credits this AI-powered surveillance in supermarkets with cutting monthly failed-transaction losses from roughly S$10,000 to S$1,500.
SCENTS N SECRETS — SCENTS N SECRETS, a Pakistani fragrance brand, launched an Agentic AI chatbot in retail service on Instagram and Facebook Messenger, tied to Shopify and Asaan Retail, an instance of Agentic virtual assistants. Week one: 4,000+ queries answered, orders taken, updates pushed live.
Supermercados TaTa — A chatty grandmother persona powers IAIA, the generative AI retail assistant VML Uruguay built for Supermercados TaTa. It chats, fields queries and suggests goods, framing an AI shopping experience around warmth rather than pure transaction.
Lenehans — Lenehans built a conversational AI shop assistant into lenehans.ie, covering over 30,000 products. Shoppers get tailored suggestions, live stock and delivery checks, plus guided DIY steps. Ireland's home retail sector rarely sees a retail AI assistant this broad.
Currys — Currys is testing its new Action AI, a retail analytics platform developed with Quorso, in 16 UK stores before it expands nationwide from September 2025. Store managers see a single dashboard combining shopper numbers, transaction levels and satisfaction ratings, powered by extensive calculations that flag which actions to prioritise on the shop floor. Retailers increasingly rely on this kind of AI-driven retail analytics to convert raw performance figures into practical staff guidance.
Seven-Eleven Japan — AI in retail operations is now company-wide at Seven-Eleven Japan: AI Library covers 8,000 employees, up from 4,000 staff in 20 departments by June 2025. Staff choose among 13 large language models, including Claude, Gemini and GPT, for minutes, product development, and POS or social media analysis. This corporate AI rollout cut minutes per meeting from 40 to 10.
JINS Holdings — JINS Holdings is piloting AI customer service retail technology through JINS AI, a self-built generative chatbot live in select eyewear stores. Shoppers scan a shelf QR code for frame suggestions, multilingual pricing and FAQ answers, part of an AI customer service system built in-house. Full rollout follows in 2026.
House of Travel — House of Travel Group launched HoT Brain, its own AI for travel consultants, bundling map creation, passport scanning, personalised itinerary generation and ChatGPT functionality. The consultant-facing AI runs in a secure Microsoft tenancy linked to Timatic and internal knowledge databases. Live in New Zealand, HoT Brain reaches Australia next month.
Lisek.App — Lisek.App, a Polish quick-commerce grocery platform, now runs AI grocery shopping through OpenAI's Agent GPT (ChatGPT's Agent Mode). Given a list, the agent finds items on the site, adds them to the basket and completes the order without manual clicks. This form of autonomous AI shopping agents works on desktop only.
Jumia Egypt — Egyptian dialect now drives AI voice shopping on Jumia Egypt, through Ziila, a voice-ordering assistant made with Intella, a firm specialising in Arabic AI. Spoken commands replace typing entirely for shoppers placing orders. Jumia Egypt's CEO opened the feature by speaking the first order himself.
Jockey Plaza — Lima's Jockey Plaza now runs MIA, an AI virtual assistant for shopping malls, built with AIFINDR to answer visitor questions instantly via WhatsApp and the mall's site. Shoppers get store directions, tailored suggestions, and event or promotion updates from the chatbot. Management calls it the region's first mall with this kind of automated visitor guidance.
Konzum — Instaleap's AI grocery platform now drives operations at Konzum, the Croatian supermarket chain, merging incoming orders from its websites, apps and marketplace channels into a single workflow. Warehouse staff receive optimised picking routes and task assignments matched to their availability, while shoppers get automatic substitute recommendations whenever stock is depleted. This AI-powered e-grocery system responds to growing consumer appetite for buying groceries online across Croatia.
BREEZEONLINE (Breeze Group) — Generative AI personalised shopping now drives BREEZEONLINE's Taiwanese luxury storefront, built on UpGPT from developer Anran Technology. Typing a plain-language wish, like wanting something elegant for an occasion, prompts the tool to surface matching picks, while behind the scenes AI image analysis auto-labels listings with style tags and suggested contexts. Conversion from search and repeat visits have both climbed post-launch, a gain the retailer has not quantified, as AI-driven product recommendations reshape how customers browse and buy.
Toolstation — Failed searches dropped 95% after Toolstation deployed Google Cloud's Vertex AI Search for Commerce, part of its shift toward AI product search in the UK. Gemini models now decode trade jargon and synonyms, driving a 5.5% rise in search revenue and a 10% jump in click-through rates. The retailer's AI search tool has since spread to its contact centre and tills.
SPAR Austria — Simbe Robotics' Tally leads Austria's first shelf-scanning robot grocery retail trial for SPAR, scanning a Eurospar in Vienna and Interspar in Eisenstadt three times daily. This autonomous inventory scanning robot flags gaps and emails staff, freeing them for customers over five months.
Shiseido Japan — Shiseido Japan's PBPs now type plain questions into a generative AI chatbot for beauty advisors, an AI-powered retail customer service tool built into their Beauty Tablet handsets, to pull up product, brand and promotion details. It cuts calls to sales and admin teams, helping staff assist shoppers faster. Shiseido Interactive Beauty and Accenture co-developed it with Shiseido Japan after a trial run.
momo (Fubon Multimedia) — Over 90% response accuracy resulted from momo's LLM-powered AI customer service platform, built with Microsoft Azure OpenAI via RAG. Self-service adoption climbed 5% at Taiwan's biggest online retailer, easing pressure on human agents. Older chatbots gave way to this generative AI support tool for natural-language shopper questions.
Magnit — Magnit's AI forecasting and replenishment system has launched at a distribution hub in Kirov, marking Russia's first fully domestic tool of this scale in offline retail. Built on open-source components such as Apache Spark and Kafka, it already covers over 600 shops nearby, using machine-learning demand prediction in retail to set stock orders by product line. The goal: cut inventory days by 10% and add tens of billions of roubles in value within a few years.
IKI Lithuania — AI facial age estimation now guards a self-checkout lane at IKI's Vilnius store on Pilaitės pr. 42, confirming shoppers buying age-restricted goods are over 25. StrongPoint built the on-device system for the Lithuania grocery chain, part of Rewe Group, shrinking a 30-second manual check to under 10 seconds and aiming to automate 60% of such checks. IKI says the automatic age recognition analysis stays on the device, sending nothing offsite and storing no images.
Fischer SA — Fischer SA, an artisanal Luxembourg bakery, now runs bakery demand prediction via a Resultance tool weighing seasonality, weather and local events. Revenue climbed 4%, costs recovered within three months, and this AI sales forecasting for bakeries debut later won the AI Adoption prize at the AI Excellence Award 2025.
Sar HaMashkaot — Ramat Gan now hosts Israel's first AI-powered store, opened by alcohol retailer Sar HaMashkaot. Its "smart room" chats with shoppers, gauges their tastes and recommends bottles unaided by staff. Founder Evitar Sdan framed the launch as a trial of automated retail AI, testing whether software alone can drive sales and suggestions.
Carrefour — Shelf by shelf, Carrefour is rolling out AI computer vision in retail stores: Edgesense units from VusionGroup, wired with networked cameras, now track products, pricing and stock levels automatically. The concept launched at a 12,500-square-metre hypermarket in Villabé, fitted with 7,000 Edgesense rails and 500 cameras. Real-time alerts flag low stock to employees, and this image recognition technology in supermarkets can point customers straight to what they're after.
Organización Soriana — Soriana, Mexico's supermarket chain, rebuilt 40-year-old mascot Julio Regalado with generative AI, the country's first retailer running an AI-powered brand influencer. It replies to TikTok shoppers instantly with personalised messages, an AI-generated promotional character, not a scripted ad.
Åhléns — AI customer service at Åhléns saw fallback queries fall from 12% to just 0.02% within weeks, after the Swedish chain launched a generative AI agent built with partner Ebbot. The virtual support agent, live since mid-March 2025, handles order tracking, delivery and returns queries on the retailer's site, built in under two months.
VARUS.UA — Manual basket assembly at VARUS.UA once took 30 minutes; with Neomi, an AI grocery shopping tool, the chain's 113 outlets now finish in 1–2 minutes. This AI recipe-based shopping feature nearly doubled conversion versus ordinary browsing, per the retailer, with over 340 online orders during Neomi's opening three-week pilot.
Albert Heijn — Albert Heijn's new AI assistant for retail staff gives 80,000 shop floor colleagues instant answers on stock levels and product locations, launched at FoodTech Week. This generative AI tool for shop floor employees is being rolled out nationwide.
Albert — Ahold Delhaize's Czech chain Albert now spots unbarcoded bread, fruit and vegetables via a small checkout-scale camera, speeding up staffed tills nationwide. The AI product recognition tool already ran in 50 outlets after a successful trial, with all 350 stores targeted by year end. This automated grocery identification push replaces manual lookups for loose produce.
Snow Factory (Xuefang Premium Yogurt) — A 30% drop in customer service spending followed Snow Factory's rollout of an AI chatbot in customer service, built on Appier's BotBonnie for LINE. Taiwan's yogurt brand feeds documents into a knowledge bot that autonomously handles FAQs, keyword triggers and tougher conversational AI customer support cases, lifting satisfaction too.
Lotte Department Store — A 70% drop in customer analysis time within one month followed Lotte Department Store's rollout of BI Agent, an AI-powered BI system layered on Strategy One, the generative platform from Strategy (formerly MicroStrategy). Wired into existing customer-data tools, it lets untrained staff pose questions in plain wording rather than code, extending AI-driven business intelligence into routine retail choices. Work once needing specialist skill, like cohort and relationship studies, now runs directly under marketers at stores and head office.
Wildberries — Wildberries brought AI in e-commerce analytics to Kazakhstan with Jam, a new seller-facing tool powered by a neural network. Sellers gain automated image and description editing for promo pages, plus visibility into audience segments, search-term performance and sales funnel conversion. The marketplace positions this AI-based seller analytics tool as a way to sharpen listings and lift conversions for Eurasian sellers.
Coop Szeged — Coop Szeged's new outlet at Szent István Square runs an AI-powered smart store model: staffed by day, then unmanned overnight via app-only entry and payment. Sensors spot odd movement without facial recognition, and a separate device identifies fruit and vegetables at checkout. Covering 150 square metres, it is Hungary's first shop blending human staff with automated retail AI.
Cervera — Sweden's Cervera deployed CAI (Cervera AI) for AI customer service retail use, closing over 80% of almost 12,000 chatbot exchanges without human staff since going live. Three in four respondents said the assistant met their needs precisely. Parent group Homeco extended this AI chatbot in retail to Norwegian siblings Kitch'n and Tilbords.
Massy Holdings Ltd — Massy Holdings runs an AI store management system on ChatGPT, letting Caribbean managers query stock, demand and sales trends instantly, cutting waste and lifting efficiency across 60-plus stores. Massy also counts over 100 self-checkout machines group-wide, part of the same AI-powered retail operations push, per its own unverified figures.
Terpel Ecuador — Terpel Ecuador now runs TBot, an AI customer service chatbot on WhatsApp Business, across its fuel stations, Altoque convenience stores and Voltex charging points. This AI-powered digital copilot replaces manual queries by sharing site locations, promotions and payment methods, plus collecting feedback.
Lowe's — Sales-floor staff at more than 1,700 Lowe's locations nationwide now use an AI retail assistant, Mylow Companion, to pull up merchandise details, project tips and warehouse counts on demand. The rollout follows Mylow, marketed as the home improvement sector's first AI-powered virtual advisor for shoppers.
OLX Ukraine — Sellers on OLX Ukraine now get an AI product listing generation tool that drafts descriptions and fills in item details from nothing more than a photo and headline, starting in the Fashion & Style category. Parameter accuracy reaches 75%, and 52% of the AI-generated item descriptions need no further editing before going live. The company credits the feature with a 40% drop in manual data entry.
SPAR Hrvatska — SPAR Hrvatska deployed AI chatbots in retail across WhatsApp and Viber, working with partner Infobip. Shoppers get nearby store locations, Sunday opening hours and weekly deals, plus live agent support and vacancy listings through WhatsApp, broadening conversational AI in retail nationwide.
Albert Heijn — Steijn, an AI kitchen assistant grocery app tool, gives Albert Heijn's Mijn AH users cooking help, recipe picks and fridge-storage tips, built on Azure AI Foundry and Azure OpenAI. Photograph a fridge and it drafts a shopping list. Roughly 5 million weekly users of the Netherlands chain's app can reach this AI recipe assistant.
Suning — Suning Max stores in Zhengzhou and Beijing now run an AI store assistant and digital human guide on Lingsi, the retailer's own large model, fused with DeepSeek. Keyword lookups return product details in under a second, closing queries in a minute instead of four hours, with accuracy up from 80% to beyond 98%. The avatar adds tailored suggestions and live deals, an early retail AI deployment gain.
Auchan Hungary — Budapest now hosts Hungary's first cashierless AI store, Auchan GO, launched by Auchan Hungary on 10 April 2025 at Bikás Park. Shoppers enter and settle bills through the OkAuchan app, while 20 overhead cameras paired with weight-sensing shelving, spread across just 18 square metres, register each item taken. Cloudpick engineered the smart store platform behind the shop, rolled out locally by partner Kende Retail Operation.
Divar — Divar, Iran's largest classifieds platform, has launched دستیار خرید, an AI shopping assistant for plain conversational buyer requests instead of fixed filters. It suggests comparisons, gives purchase advice and surfaces matching ads beyond standard search. This AI-powered shopping assistant remains a pilot, limited to property, vehicles, digital goods and certain appliances.
Mitsubishi Motor Sales of Canada — Mitsubishi Motor Sales of Canada built an AI car-buying assistant called Intelligent Companion for shoppers browsing the 2025 Outlander. Running on IBM watsonx with its Granite model, it renders 3D feature tours tailored per visitor, an AI-powered buying experience guiding them to purchase.
SPAR South Africa — Employee adoption of Microsoft 365 Copilot retail operations at SPAR South Africa has reached 67%, saving roughly 715 hours of labour. Drafting happens in Word, notes get condensed in Teams, and slides emerge from PowerPoint, weaving AI-assisted document creation into daily routines. Ninety-three per cent of staff said output had improved.
Petromin Corporation — Close to half of Petromin Corporation's WhatsApp customers tried PETROMINit!, its AI customer service agent, in its launch month, mainly for quick vehicle servicing and loyalty points. Built on large language models with conversational AI platform Gupshup, the Saudi mobility firm's automotive customer service chatbot replies in Arabic and English and also covers urgent breakdown help.
Fnac Darty — Fnac Darty runs about ten GenAI use cases in production for generative AI in e-commerce: catalogue enrichment, search, recommendations, review summarisation and support automation. The French retailer enriches listings across 15 million offers reaching 27 million monthly unique visitors, adding conversational search plus AI-powered customer service for its top ten interaction types.
Al-Futtaim — Over five million Blue Rewards members can now use Blue AI, Al-Futtaim's embedded retail AI assistant suggesting fashion, beauty and other items via single-tap multi-brand checkout. Debuting in the UAE, this AI shopping companion should extend to eight additional markets during its first year running.
BIPA — BIPA staff in Austria now use Haidi, an AI assistant built with Google Cloud, rolled out chain-wide after limited trials. REWE Group's drugstore chain deploys this AI store assistant for queries on returns, complaints and workflows in German, English, Turkish and Hungarian. This virtual employee helper expanded once testing succeeded.
Carrefour Belgium — Carrefour Belgium runs a 24/7 checkout-free AI retail store at the Nato SHAPE base near Mons, Casteau — its biggest BuyBye micro-store yet, third overall, with 12 smart entrances. Built with Reckon.ai, cameras and weight sensors spot what shoppers take and charge them on exit, no tills needed. Over 250 daily goods fill the shelves, growing Carrefour's cashier-less shopping network.
Kjær & Sommerfeldt — Kjær & Sommerfeldt, Denmark's oldest wine merchant, built an AI wine advisor pairing bottles and answering shopper queries across roughly 2,000 wines with expert-level guidance. It took over a year and up to 20 million Danish kroner, partly spent cleaning the catalogue's underlying data. An early case of AI-driven product recommendation in wine retail.
Megasuper — A first for Central America and the Caribbean: Toshiba's ELERA Produce Recognition now runs inside Megasuper stores in Costa Rica, bringing AI produce recognition in retail checkouts to the region. Lenses catch unbarcoded produce mid-scan, freeing shoppers from typing codes. The rollout extends Toshiba's MxP Self-Checkout kiosks, already beyond 10% of Megasuper's lanes — computer vision item identification at self-checkout, now part of daily grocery runs.
Carrefour — Retail AI cataloguing underpins Carrefour's inventory of 850,000 active products sourced from 23,000 suppliers, joining Google Gemini's GenAI with OCR scanning of packaging photos to populate listings unattended. This AI product catalogue enrichment achieves a 78% accuracy benchmark, free of errors, for branded merchandise. Elsewhere, a trial identified and resolved 3,400 supplier mistakes spanning 108 suppliers.
Carrefour LabelVie (LabelVie Group) — Carrefour LabelVie now runs the LEAFIO Shelf Efficiency system to handle AI planogram management across 180 stores in Morocco, spanning four formats from convenience shops to hypermarkets. Store teams execute layouts on mobile tools instead of building shelf plans by hand, since the platform generates planograms automatically and drives merchandising through data. The rollout, described as complete, is being extended across the retailer's full multi-format network through automated shelf planning.
Euronics Italia — Ronics, Euronics Italia's AI shopping agent on Salesforce Agentforce, logged over 8,000 conversations, lifting conversion 53% and drawing a 7% click-through on product picks. It brings physical-store know-how to the retailer's website, part of Italy's shift toward AI-powered e-commerce recommendation systems.
MC (Sonae Group) — Sensei's AI computer vision and sensors power MC's 1,200-square-metre Leiria outlet, billed by Sonae's retail arm as the world's biggest autonomous supermarket. Overhead lenses and shelf sensors track shoppers' picks, cutting scanning and till queues, trialling checkout-free retail technology in Portugal.
Pingo Doce — Symphony AI built the AI merchandising platform now running at Pingo Doce, giving the Jerónimo Martins-owned Portuguese chain a shared system for CPG partners. Shelf, assortment and till data sit in one layer, letting staff model outcomes and track performance live for retail category management.
Delhaize Serbia (Maxi) — MaxiGPT, an AI digital assistant for store associates, now runs across Delhaize Serbia's Maxi, Mega Maxi, Tempo and Shop&Go outlets, handing shop-floor teams fast answers on stock and product data. Ahold Delhaize revealed the tool's launch within second-quarter 2025 results, framing it as a productivity aid for the country's biggest supermarket chain. The broader push toward an AI store assistant for retail staff reflects efforts to streamline everyday branch operations.
Epitsentr K — AI customer support in retail now resolves a third of Epitsentr K's queries through an AI-driven customer service tool on Facebook, Instagram, Telegram and Viber. More than 50 self-service tools manage orders, payments and shipments without staff involvement. Ukraine's retailer aims for 70% automated coverage by 2026.
KIDO Group — KIDO Group's Mid-Autumn 2025 sales tripled from a year earlier once it turned to AI livestream sales and machine-generated marketing content across the business. A digital avatar hosted a broadcast lasting beyond 250 unbroken hours, drawing 19 million viewers and shifting more than 180,000 items while a single staffer kept watch over the feed. TV ad spots emerged 60% quicker and at almost 40% less expense thanks to the same push into machine-generated marketing content. Afterward the firm set up AIRO, a separate AI arm handling sales and marketing, on 30 December 2025.
Designer Wardrobe — Designer Wardrobe's DW AI List Assist trims listing time by up to 90 per cent for 350,000-plus members of New Zealand's biggest pre-loved fashion marketplace. From snapped pictures, the AI listing tool completes each field within seconds. AI-drafted listings suggest live prices from recent site sales.
Healthspan — Healthspan, a United Kingdom vitamins and supplements retailer, resolved 90% of product queries within eight weeks using a generative AI chatbot from Talkative in its contact centre. The AI customer service chatbot lifted satisfaction from 86% to 89%. With automated support absorbing routine questions, Healthspan's 57-person team handles higher-value work.
OWNDAYS — 49 OWNDAYS branches across Japan have adopted JetB's AI avatar customer service system, expanding beyond a six-month trial at two Tokyo and Osaka shops. Branded 'Uchi no AI Avatar', it offers multilingual customer support via in-store QR code, conversing in 15 languages and reaching native speakers among over 90% of the global population. Staff can inspect logged chats before stepping in.
Packworks — Filipino sari-sari shops running Philippines-based Packworks' AI-driven retail analytics tool, Store Insighting Project (SIP), saw daily gross merchandise value climb 46% and overall revenue rise 17% inside a fortnight, across a network topping 300 outlets. Despite trading on 20% fewer days, participating vendors still out-earned peers, evidence that AI-guided inventory and demand planning beat simply staying open longer.
Visa — Moroccan shoppers rely on AI in online shopping at a rate of 83%, per Visa's survey with Bank Al-Maghrib and Switch Al Maghrib — a pattern of AI-assisted e-commerce seen in gift-hunting (57%), price checks (55%) and reading reviews (42%). Trusting an autonomous AI agent with an actual purchase, though, draws only 23%.
Visa — Visa's Stay Secure study finds 91 per cent of consumers in Egypt have used AI tools for shopping, for price comparison (75 per cent), gift ideas (65 per cent) and reviews (65 per cent). Wakefield Research's annual poll of 5,800 adults across 17 CEMEA markets shows 97 per cent think AI-powered tools make online buying quicker and simpler, and 88 per cent expect AI to be critical to future fraud protection. AI shopping assistance stops short of payment: only 38 per cent currently trust AI agents with checkout.
Visa — Visa's Stay Secure study by Wakefield Research found 89 percent of Kenyan consumers used AI shopping tools: 60 percent checking reviews, 55 percent finding gift ideas, 53 percent comparing prices. Though 91 percent say AI-powered tools speed up and simplify buying online, only 29 percent would let AI agents complete checkout. Sample: 5,800 adults across 17 CEMEA markets.
Cavecom-e — Cavecom-e, the Venezuelan Chamber of Electronic Commerce, put the share of active online retailers using AI in e-commerce at 80%, mainly to lift sales and follow up with buyers before and after purchase. Richard Ujueta added that 40% overall are starting to add AI agents for social-commerce queries.
Adyen — More than half of Brazil's consumers, 52%, leaned on an AI shopping assistant such as ChatGPT during the past 12 months, Adyen's Retail Report 2025 reveals, a figure up 52% from the prior year. Of that share, 74% credited the assistant with steering what they bought when shopping with AI. Researchers polled 2,000 consumers alongside 500 retailers nationwide.
Adyen — More than a third of Danish shoppers—35%, per Adyen's 2025 Retail Report spanning 41,000-plus consumers across 28 countries—now turn to AI in retail shopping, a 46% jump on 2024. Younger buyers dominate: Generation Z tops the list at 56% uptake, whereas Generation X posted the biggest climb, rising 54%. Nearly half of Denmark's retailers, 43%, expect to fund AI-driven sales and marketing within the coming 12 months.
National Consumer Service (Servicio Nacional del Consumidor - SERNAC) — SERNAC, Chile's consumer protection agency, addressed AI consumer protection via Resolución Exenta N° 33, guiding businesses on privacy hazards from automated tools used with shoppers. Expectations cover algorithmic disclosure, fair treatment, secure data handling, banned dark patterns, and human review of automated decision-making affecting consumers.
Post Bellum — Post Bellum has thrown open the record of its Paměť národa collection: some 30,000 hours of survivor testimony, machine-transcribed and free to interrogate through a new AI assistant for oral history. Put a plain query rather than chase keywords, and the tool locates the passages that bear on it. Landing as Post Bellum and Paměť národa reach 25 years, the release carries the director's caution that no witness will be falsified by AI-assisted archive search.
Hubei Provincial Museum — Hubei Provincial Museum made ByteDance's Doubao its official AI museum guide, speaking 12 Chinese dialects including Cantonese, Hokkien, Shanghainese, Hakka and Wuhan. Visitors press a guide button, aim a phone at an object and get a tailored explanation. The virtual docent answers follow-ups by video call.
ETV — ETV has fronted its bulletins with AI news presenters for over a week, adding on-screen captions that tell viewers which footage a machine made or altered. Cindy Martin delivers Guadeloupe bulletins and flashes; Jean Louis handles Caribbean coverage. Clips carry the caption "IA — GÉNÉRÉ PAR IA". Synthetic anchors also compress the production chain, which the broadcaster credits for six different editions daily.
MTN South Africa — MTN South Africa has set telecom AI fraud detection onto behaviour patterns to pinpoint likely data tunnelling, chief executive Ferdi Moolman says. Tunnelling dresses billable traffic up as visits to zero-rated sites; network traffic screening raises anomalies for investigators, sparing analysts a manual sweep of each exempted destination.
KSAT 12 — KSAT 12 in San Antonio has launched The Rundown, a weekday 1 p.m. half-hour newscast without studio anchors, where AI-voiced teases link some stories. Newsroom staff approve each synthetic voice clip before air; reporters front stories from the field, with a live weather anchor.
KT — KT now runs AI stadium operations daily at Suwon KT Wiz Park, from facial-recognition entry and synthetic-voice announcements to motion-recognition fan cameras and live translation for foreign spectators. KT says entry, previously 50 to 60 seconds a person with bag checks, is about 30 seconds faster, more than halving delays. Foreign spectators' attendance share rose 57% in the first half against a year earlier after translation launched.
T-Mobile Czech Republic — T-Mobile has cut AI customer support resolution times in Czechia by up to 28 minutes per live-chat request, using the Sprinklr platform and its AI assistant, versus roughly 80 first-quarter users. Over 500 staff now access the tool, which surfaces information and drafts replies. This marks a scale-up in AI-assisted customer service, though the savings figures are not independently verified.
MTN Nigeria — MTN Nigeria launched Zigi Voice, an AI voice assistant letting customers talk live in the myMTN NG app. Requests include NIN enquiries, data purchases, balance checks and personalised recommendations. Spoken AI support runs in English and Pidgin, with more local languages being developed.
Broadcast Solutions — Broadcast Solutions produced the inaugural Franz Beckenbauer Supercup's tactical feed with automated camera production systems from Studio Automated, plus Egripment equipment. The company says the systems delivered the tactical coverage but gave no detail on the automation behind it.
The Washington Post — The Washington Post has built AI agents on ChatGPT and OpenAI's APIs that answer recurring advertising, subscription and content questions from its internal datasets, within permission rules. Cross-dataset lookups become an on-demand assistant, leaving analysts free for judgement, per a trade-news summary with no scale figures or named sources.
WPP & Omnicom — WPP and Omnicom are the first agency partners for OpenAI's ChatGPT Ads in India, covering strategy, creative work, campaign execution, measurement and brand safety. OpenAI says the deal is non-exclusive, that over 50 brands planned to advertise on ChatGPT there in launch week, and that self-serve access follows in early September.
Verizon — Verizon has deployed Google Cloud's Gemini AI across marketing, network operations and customer service. In some channels Gemini handles most inbound consumer calls and chats, diagnosing issues and passing complex cases to staff; AI agents patch and re-route network systems automatically via programmatic access. Financial terms, timelines and figures went undisclosed.
International Press Centre — The International Press Centre piloted WAFLO, a WhatsApp-based AI election incident reporting system, during the Osun State governorship election. Journalists on duty logged harassment, intimidation, assaults, threats and other press freedom violations live over the WhatsApp line for monitoring and documentation. It runs under Component 4 of the European Union Support to Democratic Governance in Nigeria Phase Two programme, with the Centre for Media and Society.
Telefónica — Telefónica has built generative AI into its business voice services in Spain as 'Centrex AI', across corporate mobile lines and Centrex IP and Centrex 365 virtual PBX products. Features include speaker-labelled, near real-time call transcripts, AI summaries and virtual agents that check records and book appointments, in 100-plus languages. Spanish firms and public administrations come first.
TV Asahi — TV Asahi has aired Japan's first AI-generated TV commercial: a 30-second spot for Suntory's GREEN DA-KA-RA drink, made wholly from machine-generated visuals. Its AI Creative Studio, formed April 2026, steers generative AI advertising across animation, drama and promotional output.
Bell — Bell has launched AI-based 'Suspicious Call Detection', which automatically spots and blocks spoofed calls on iOS and Android for customers of Bell, Virgin Plus, Lucky Mobile and its other mobile brands, with no opt-in. Bell calls itself Canada's first carrier to flag spoofing with AI; its existing anti-fraud technology had analysed 4.4 billion calls and blocked or labelled over 540 million suspicious ones.
TelevisaUnivision — TelevisaUnivision confirms it already uses AI dubbing on part of its catalogue to take productions faster into markets where Spanish is not the main language. It works with ElevenLabs and other specialists whose voice replication translates into Portuguese, Korean and more far quicker than conventional dubbing.
Telekom Srbija — Telekom Srbija, Serbia's leading telecoms operator, has rolled out SAS Customer Intelligence 360, SAS Intelligent Decisioning and SAS Viya to unify AI in telecom marketing for over 8 million customers. Nine previously separate teams are now merged into one system, enabling live personalisation and AI-driven decisioning across mobile, web and self-service touchpoints.
Annahar Media Group — Annahar Media Group aired a bulletin read entirely by AI news anchors in Beirut on 27 July 2026. Digital presenter Noor interacted directly with digital correspondent Nahar, billed as a first for Lebanon and the Arab world. Points Information Technology, a Lebanese firm led by Bilal Kassassir, runs the strategy.
BusinessDay — BusinessDay has unveiled Nigeria's first AI news assistant, SIA (Strategic Intelligence Assistant), becoming the first newspaper AI chatbot in the country. Readers pose questions in ordinary language and get responses pulled straight from the paper's own articles, with the tool condensing dense subjects and locating pertinent pieces across markets, policy and companies. No other Nigerian title had rolled out such a tool before.
Tigo (Colombia Móvil and UNE EPM Telecomunicaciones) — Colombia's Communications Regulation Commission (CRC) has opened the testing phase of ConTIGO Digital, Tigo's project for AI customer service by phone. Full Tigo subscribers are first to try the phone AI, linked to the operator's internal systems, and the CRC will use the results to weigh regulatory changes.
Moov Africa Malitel — Moov Africa Malitel, Mali's biggest telecom carrier, unveiled an AI voice assistant called MIA (Moov Intelligence Artificielle) on 12 June 2026, reachable by dialling 35201 even from basic handsets lacking internet. Callers get instant guidance on farming, schooling, wellbeing and daily needs, billed from 25 FCFA per minute. The voice-based AI assistant push targets broader digital access nationwide.
MTN Nigeria — MTN Nigeria runs 13 AI agents for SIM registration verification, work more than 200 workers did by hand. Using computer vision, the agents beat manual processing on speed and accuracy. MTN is extending automated SIM checks to Ghana, Cameroon and Zambia, among other African markets.
BelTA — BelTA has built AI photo restoration and video work into daily production, director general Andrei Mokhor told the 34th Council of Heads of CIS State News Agencies in St Petersburg. The Belarusian Telegraph Agency animates archival and historical photographs, produces colourised archive footage, generates billboards, posters and infographics, and upscales old frames for large-format exhibitions. Writing or editing copy stays off limits — Mokhor calls that a taboo.
485 Center for Generative Technologies — Little Mher, an 80-minute animated feature drawn from the Armenian epic Daredevils of Sassoun, reached Yerevan screens on 31 May 2026 and went on general release across the country. The 485 Center for Generative Technologies confined AI film animation to drawing frames and moving them: script, vocals, score and art styling stayed with a crew of actors, musicians, screenwriters, illustrators, cinematographers and machine-learning experts, according to director-producer Hovhannes Margaryan. Matching machine-made pictures to Armenian dialogue proved hardest, since such models mostly learn from English. Its makers bill it a regional first.
Qatar News Agency — Qatar News Agency has added AI-produced news services in Russian, Hindi and Urdu, lifting its publishing languages to eight with Arabic, English, French, German and Spanish. The director-general says AI had already been part of the agency's news system for a while.
Kriol Think — Kriol Think fielded 802 queries over a 14-day run at 100% uptime through the AI electoral assistant Nós AI Eleitor 2026, built by Edson da Veiga for Cabo Verde's legislative elections. Anthropic's Claude Haiku 4.5 drafted every reply strictly from the parties' own electoral programmes, three loaded, MpD, PAICV and UCID, while PTS and PP awaited release, and each response pinpointed the underlying passages. Guardrails barred political views, party endorsements and fabricated material, obliged the assistant to concede gaps, and excluded identifiable personal data, stored query wording and tracking cookies. Unaligned and free, this election AI tool emerged from a non-profit civic venture whose running costs Edson da Veiga chased via GoFundMe; the usage totals are Kriol Think's own, surfacing on an archive page after the ballot.
Punktum.fo — Heini Festirstein and Eyðstein Johannesen have launched Punktum.fo, a Faroe Islands news site built on AI news writing. Punktum.fo pulls stories from vetted sources into short, plain-language Faroese summaries, each with an audio option. Every source stays visible on the page. The pair made it partly to push Faroese-language AI forward, calling that transparency non-negotiable for AI-generated news.
Ministry of Information, Culture and Tourism (Somalia) — Somalia's Ministry of Information, Culture and Tourism now uses AI translation for government media, folding automated content generation for public broadcasting into one digital distribution pipeline. A ministry communications consultant told the East Africa Science and Technology Commission summit the shift has cut editorial production time sharply, replacing manual steps with a faster information network.
Radio Ndeke Luka — Radio Ndeke Luka's Bangui verification desk, #StopAtenè, ran AI deepfake detection, the tool Hive Detection, against a clip alleging Kim Jong-un's pledge of North Korean military backing for Iran. It measured a 98.5% probability the clip's voice was synthetic, with manipulation markers elsewhere in the footage, and separately caught the clip misdating a Kim address to 25 March 2026 against his genuine 23 March appearance. #StopAtenè's verdict, published 2 April 2026 through AI-generated video verification, found no proof behind the claim.
Philippe Shangti — Philippe Shangti unveiled Remember the Future: 2320 at Cinemes Illa Carlemany in Andorra la Vella, a 95-minute sci-fi title billed as the nation's earliest AI-generated film. Some 45,000 visual prompts fed the assembly across almost twelve months. Camera crews sat out this AI filmmaking.
Jogorku Kenesh (Parliament of Kyrgyzstan) — Kyrgyzstan's Jogorku Kenesh has unveiled AI news anchors, Myrza and Aiyym, reporting on legislative business. These virtual news presenters turn out written summaries and clips explaining debates, keeping citizens informed of deputies' activities. Speaker Marlen Mamataliev tied the launch to the legislature's digitalisation push.
NTT DOCOMO — Since 4 February 2026, Japan's mobile network has run agentic AI network maintenance under NTT DOCOMO, with the platform scanning alarm and traffic signals from more than one million connected devices to surface anomalies and propose fixes automatically. Engineers now resolve complex faults over 50% faster, per the carrier's own measurement. Built on one of the world's largest telecom datasets, the deployment marks a shift toward AI-driven fault detection at scale.
Fabric of St Peter — The Fabric of St Peter brings AI simultaneous translation to St Peter's Basilica, carrying liturgies live in up to 60 languages, partnered by the Dicastery for Communication and the technology company Translated. Cardinal Mauro Gambetti, archpriest of the basilica, unveiled the scheme four centuries after the 1626 consecration. A QR code inside the church opens a browser page relaying readings, prayers and chants as sound and text in each pilgrim's tongue, nothing to download. Antonio Autorino, communications head there, dates full running of the machine interpreting to spring 2026, trials among chosen worshippers having already begun.
BBC — The BBC has extended Style Assist, an AI in journalism tool, to further UK newsrooms to convert Local Democracy Reporter Scheme copy into BBC style and tone. A senior journalist checks every piece pre-publication. My Club Daily, an expanding newsroom AI pilot, sends personalised AI-generated audio bulletins to football fans.
Tele2 — Mobile operator Tele2 has rolled out AI customer call analysis from Latvian startup asya.ai for customer service. Pitch Patterns, an AI conversation analytics tool, reads tone, intonation and speech pace in Latvian and other languages, flagging unsuccessful interactions to improve service quality without manual listening.
Annahar Media Group — Annahar Media Group has equipped its newsroom in Lebanon with an AI news correspondent, Nahar, assembled with Points Information Technology. Nahar sweeps thousands of outlets within moments, follows world affairs, weighs rival accounts, renders languages and condenses documents. Editor-in-chief Nayla Tueni cast the AI reporter as a colleague enlarging what journalists achieve.
Centrafrique Check — Centrafrique Check, a Bangui fact-checking outlet, ran AI image detection with two tools, MyDetector and TruthScan, on a photo attached to claims a French national had been arrested in Zémio, Haut-Mbomou, circulating from around 7 March 2026. Médecins Sans Frontières had confirmed the detention, around 4 March 2026, of two of its own staff, one French and one Central African, taken to Bangui. Both tools put roughly 98% odds on artificial generation, pointing to a skin-tone mismatch on the soldier's hand where it gripped the detainee. AI-generated photo verification found the arrest genuine, the picture false.
Television Cubana — Television Cubana has fielded two AI-generated presenters, Ava and Chico, across trailers for the music show Cuerda Viva, appearing as announcers, or reading round-ups and listings. Cari Rojas, communication director of Television Cubana, dates Ava to a Cuerda Viva festival; Ana Rabasa, who runs the show, brought in Chico later. Felito Ruiz, the team's designer, assembled both on no-cost software of limited scope, and calls these virtual TV hosts wholly invented looks, resting on no prior pattern.
Curaçao Innovation Hub — Curaçao Innovation Hub placed second worldwide at the SparkVision AI Film Festival 2026, a contest open only to work made 100 percent with artificial intelligence. The Ministry of Economic Development (MEO) released the placing, casting the hub's AI filmmaking as what emerging technology and innovation yield beside Curaçao's culture, language, identity and local talent, and as proof the island can compete internationally in fast-moving technological and creative industries. Makers of the AI-generated short film say the point was to let people see and feel a possible island future, one now waiting to be built.
RTL Hungary — RTL Hungary assembled the opening of its 'A legjobb ajánlat' second-season intro through AI video production, wrapping it inside half a day. Creative director Verebély Marcell put the choice down to a David Lynch-styled promo whose surrounding imagery resisted conventional filming; the team worked it up during April. Generative editing tools also salvaged murky, washed-out January footage.
Orange Liberia — Orange made Lorpu AI reachable by basic call or SMS, skipping smartphones and data plans entirely. The AI voice assistant Liberia now hosts spoken lessons and live commodity prices for students and small business owners, priced at L$0.5 once LAD is texted to 2026.
Grupo OPSA — An AI chatbot for election disinformation, Chekeabot, went live from Grupo OPSA's EH Verifica and LA PRENSA Verifica, resolving voter queries on ballot procedures during Honduras's 2025 elections. Staff wrote and vetted every answer rather than letting the system run unchecked. This reflects the group's push toward AI-assisted fact-checking alongside standard newsroom practice.
TRT Avaz — "Oyun Avcıları" (Game Hunters) debuts on TRT Avaz as Turkey's first AI animation in television production, an AI-generated animated series of 13 parts. Makers combined 18,000 visual assets and 60,000 seconds of machine footage, merging over 10 distinct AI systems into one pipeline.
LG Uplus — LG Uplus released Agentic Callbot, South Korea's first subscription-based AI customer service callbot, using OpenAI's generative AI. Instead of pre-programmed scenarios, agentic AI reads customer intent and conversational context to resolve complex inquiries. The automated call handling links to applications in manufacturing, distribution, hospitals and universities.
Dish Media Network (Dish Home) — Khusi, an AI customer service agent from Dish Media Network, parent of ISP Dish Home, now answers subscriber queries in fluent Nepali. This AI customer support agent gauges caller mood, catches technical faults unaided, and rings lapsed accounts to urge renewal. It debuted on Facebook Messenger and webchat.
Briefly News — Briefly News nearly doubled output with "Editorial Eye", an AI in journalism tool for proofreading and grammar checks. Daily articles rose from 80–90 to 150–200; page views grew 22% in six months, via Microsoft's Thomson Reuters Foundation editorial software programme.
One New Zealand — Coverage complaints get resolved 80% faster at One New Zealand via Network Concierge, an AI agent network diagnostics telecom system on Amazon Bedrock. It merges alarm feeds, incident history and performance stats automatically, enabling this AI-powered network troubleshooting. Storms trigger another use: ranking cell towers for backup generators.
Tobinco Media Group — Bullet TV, under Tobinco Media Group, now airs Ghana's first AI newscaster, Margaret Tobbin. The AI news anchor fills in when staff work remotely or step away, keeping bulletins on schedule. Her debut came in November 2025, timed to the station's founding anniversary.
Ooredoo Maldives — Instant AI eKYC (electronic Know Your Customer) now activates SIM cards across Ooredoo Maldives' digital channels, trimming registration to 10 seconds from a previous 60, with automatic sign-off for over 90 percent of applicants via this AI-driven identity verification. A tiered screening pipeline drives the gain, combining facial recognition, liveness detection and anti-spoofing safeguards to confirm each customer. Physical paperwork and in-person queues disappear entirely, so resort guests and locals alike can complete SIM registration straight from a handset.
SPH Media — Berita Harian and Tamil Murasu now feed articles through AI voice cloning for news, replacing generic narration with synthetic voice narration built from replicas of each newsroom's own reporters. SPH Media built the tool, powered by BeyondWords, straight into both titles' apps and sites, giving Singapore's Malay- and Tamil-speaking readers a recognisable voice rather than a robotic one. Playback has climbed to five times the publisher's normal benchmarks, suggesting audio built on real staff voices lands better than standard text-to-speech.
Diario al Día (Santos Media Group) — Manual newsroom effort has been replaced by seconds-long automation at Diario al Día, a Dominican outlet: ChatGPT now handles data verification, source checking, grammar fixes and copy prep in a fresh AI in journalism setup. Clip-making sped up too, dropping from a prior 20 to 30 minutes to mere seconds through combined use of ChatGPT, HeyGen and ElevenLabs — an example of generative newsroom tools taking over hand-done work.
Hamrah-e Aval (MCI) — IVA (Interactive Voice Assistant) handles MCI calls as Iran's first fully Persian AI voice assistant for telecom customer service, running 24 hours a day, seven days a week via Persian natural language processing. Hamrah-e Aval, the biggest carrier, folds this AI-powered telephone customer service assistant into digital transformation.
K-Towers — Kacific subsidiary K-Towers has fitted 64 remote towers in Papua New Guinea's Western Province with PowerX for AI tower maintenance. It reads power and telemetry data to catch faults alone and rank which sites need attention, ending manual checks that often needed helicopter trips. Automated infrastructure oversight now gives crews instant tower visibility.
International Centre for Investigative Reporting (ICIR) — Native AI gives Nigerian journalists their first homegrown AI transcription and translation tool, built by the International Centre for Investigative Reporting. The free system converts English audio into written Hausa, Igbo and Yoruba, easing workloads for students and researchers too, as a language-conversion aid tailored to Nigerian tongues.
CNN Prima News (Skupina Prima) — CNN Prima News debuted Nora, an AI news presenter, on AI News, a video podcast airing every week from 1 November 2025, co-created with the Czech AI Association and YDEAL. Beside human co-host Laura Doubková, the digital anchor showcases everyday AI software.
DR (Danmarks Radio) — DR built Mitvalg, an AI personalised election news tool ranking Danish local stories by relevance for the 2025 municipal and regional elections. Algorithms comb 63 hours of daily P4 radio across nine districts, breaking broadcasts into items labelled by place, topic and subject. Editors verify each piece before it goes live at dr.dk/mitvalg, so this AI-driven local news personalisation stays supervised.
SoftBank Corp. — Mobile network operations at SoftBank now run partly on its Large Telecom Model (LTM), a generative AI foundation model for telecom operations built around the company's own Sarashina large language model and hosted inside Japan-based data centres it controls. Forecasting accuracy topped 90 per cent against real readings when the system was tested during a fireworks festival, matching predicted signal quality to actual outcomes. A later upgrade added reinforcement learning, letting the model retune base-station settings on its own through an autonomous feedback loop for AI in mobile network management.
News UK — Minutes, not days: that's the new turnaround for radio commercials at News UK, whose Octave AI system now drives automated audio ad creation from concept to finished file. Made with AudioStack, the tool blends synthetic voices, scored soundtracks and sound effects into spots cleared for broadcast or podcast placement. A Dynamic Creative Optimisation function then tailors each version to whoever's actually listening, an approach to AI-generated radio advertising already being trialled by launch partners Marks & Spencer and WPP Media.
Channel 4 — Kalel Productions gave Channel 4's Dispatches documentary Will AI Take My Job? an AI presenter television had never broadcast before on British screens. Seraphinne Vallora crafted the AI news anchor's face, voice and gestures purely by algorithm, keeping viewers unaware until a late reveal.
Boomerang Stream — Simón Bolívar returns as an AI-recreated podcast host on Boomerang Stream's "Un Casacho con Simón", pulling one million views within seven days. Actor Luis Bredow lends the performance, blended with digital post-production so this virtual interviewer built with AI can question guests from Bolivia's culinary, cultural and film scenes.
DR (Danmarks Radio) — DR Nyheder reporters can scan 77,000 articles spanning five years by meaning, surfacing 45,395 possible expert voices even without exact wording matches. This AI source archive for journalists swaps out an older keyword-only directory resembling a digital phonebook, using a semantic search tool for finding expert sources built into a Microsoft platform under a formal data processing agreement.
ITV — ITV's new AI ad creation for SMEs tool, GenAI Ads Manager, lets small businesses build a TV commercial from photos and logos already sitting on their phone. Powered by Magnite's streamr.ai engine, it turns those digital assets into a finished spot in under thirty seconds, checking each one against the UK Code of Broadcast Advertising as it goes. ITV has trained its entire commercial team on the system, part of a wider push toward AI-generated TV advertising.
Globe Telecom — Globe Business, the enterprise arm of Globe Telecom, runs AI customer service automation on Salesforce's Service Cloud, Sales Cloud, Einstein for Service and piloted Agentforce, cutting workload 34 percent. Duplicate tickets vanished, trimming volume 28 percent; intent tagging scores 80 percent accuracy. This Philippines telecom's AI-powered customer support shift proves efficiency gains.
MEO (Altice Portugal) — Two AI systems have entered advanced deployment at MEO, reflecting broader AI in customer service across Portugal. An automated tool speeds fault resolution for field crews, while call centre staff use software that handles enquiries faster. Shown at Tech Trends 2025 in Funchal, these AI customer support tools sit beside a billing AI still under construction.
Banglalink — Bangladesh gets its first telecom AI-powered complaint management system, launched by Banglalink. Handset, SIM and network readings feed the tool, which flags faults and drafts self-diagnosis reports largely without staff, escalating unsolved cases to a live agent. The launch adds an AI-driven diagnostic layer to the country's telecom customer support.
ORF — Deaf viewers in Austria gain AI live subtitling as ORF flips the switch for Bundesland heute regional news and ORF Sport+, streamable via ORF ON or teletext page 777. Roland Weissmann frames this automated captioning debut as answering a long-held demand and a push toward covering 100 percent of all broadcast output.
Delfi Meedia — Delfi Meedia runs AI newsroom automation across 12 brands, an example of AI-powered news production tools: a live translation feed linking Estonia, Latvia and Lithuania, a 25-year searchable archive, court filing summaries pushed via Teams, and an advertising chatbot. Ivar Krustok said eight years built the unified data store underpinning it all.
Show TV (Veliaht) — Ercan Kesal looks 20 years younger on screen after Mojo FX ran an AI de-ageing actor process for his part as Zülfikar Karslı in Show TV's Turkish drama Veliaht. Serkan Semiz called it Turkey's first use of this digital rejuvenation method in television production. No recasting needed; software altered his appearance directly.
CJ Onstyle — CJ Onstyle's live shopping chatbot AiON, co-built with AWS's GenAI Innovation Centre, tripled reply speed and scored 85.7% satisfaction in a 'Transit Beauty' mobile broadcast trial. The live commerce assistant sorts viewer messages into five types — delivery, product, authentication and order — answering in under 10 seconds on average. Six top channels received it first.
Bouygues Telecom — Over 5,000 staff joined Bouygues Telecom's agentic AI platform, Studio IA, within weeks, building 800-plus AI agents, 51 at scale in the catalogue. Built on Prisme.ai, opened company-wide, it shows enterprise AI agent adoption scaling fast once self-service tools reach everyone.
ABC — Frame-level searches across decades of radio, TV and digital archives are now possible for reporters using ABC Assist, a multimodal AI assistant for journalists built internally at ABC. Roughly 600 to 700 employees had already begun relying on this AI research tool in newsrooms, ahead of a wider push to reach the broadcaster's full 5,000-person headcount.
Urug.mn — Soningoo (Өрөг овогтой Сонингоо), Mongolia's first AI news commentator, now fronts Urug.mn's weekly Saturday recap of major headlines and emerging trends. The virtual presenter lets the outlet trim staffing costs without sacrificing professional, ethical coverage, reflecting a push toward automation and fresh technology in local journalism.
BH Telecom — BH Telecom's new generative AI virtual assistant, "Moj Asistent" (My Assistant), now fields fault reports, account queries and package picks around the clock, launched to mark the telecom's founding anniversary. Web and social channels reach it 24/7, where it personalises answers and troubleshoots technical issues, reflecting wider telecom adoption of AI-driven virtual assistants.
O2 Czech Republic — Alenka, anti-scam AI in telecoms from Czech operator O2, poses as an elderly woman, wasting fraudsters' time and blocking up to 100,000 scam calls daily. Built on Deepgram and ChatGPT, this conversational AI fraud prevention tool keeps criminals talking so real victims never pick up. O2 seeded her number into scammer lists over summer before the rollout.
TV Markíza — Recreating wartime Bratislava for the 11th season of historical drama Dunaj, k vašim službám, TV Markíza turned to AI video generation, transforming archival photographs into convincing Second World War bombing footage. Producer Peter Kelíšek said this marked the broadcaster's first collaboration with the technology, and he found it remarkable how believable the AI-generated imagery looked when built from nothing more than old photographs. The series premiered in August 2025.
Safaricom — Kenya's Safaricom and Huawei's CBS Billing team have rolled out 'Idea to Cash', using large language models for AI in telecoms offer design, configuration and marketing. Since July 2025 a cross-functional group has applied telecom AI to hyper-personalised campaigns, offer concepts and tariff configuration from natural-language requirements.
Zain Kuwait — Kuwait hosts the region's first AI assistant for site engineers: Zain Kuwait's FME Mate, built with Huawei, is now live network-wide. The tool mines large language models to flag fault causes and recommend fixes for wireless upkeep. Zain's figures show efficiency above 20 percent in about 90 percent of cases, cutting projected yearly downtime past 2,250 hours — unverified independently.
Discover Pakistan — Discover Pakistan built an AI news channel called Pakistan Today English News, staffing it entirely with synthetic anchors, reporters and analysts instead of human presenters. Machine learning and language-processing systems animate each figure through more than 1,500 facial movements, enabling non-stop 24/7 output from a London base transmitted by satellite across the globe. Marketers describe this synthetic-broadcast approach as the first fully AI-powered English-language channel anywhere.
Telkom — Telkom's PoliPal is an AI-powered HR chatbot letting staff search internal company policies directly. It joins R46 million the South African telecoms group spent on AI in the 2025 financial year, funding Bafo, an AI contact-centre platform, and Voice of the Customer analytics.
TVB — TVB became the first Hong Kong broadcaster to stream an AI-generated short drama, "You Are The Only One In My Heart", on myTV SUPER and TVB Plus from 15 September 2025. Episodes stay under two minutes, made with AI video tools, extending TVB's earlier AI beauty pageant.
Mediacorp — Mediacorp's AI video production suite compresses prime-time bulletins into digital clips within minutes, near 90 per cent accuracy, replacing a sixty-minute manual routine. Visualiser GPT drafts narrative structure, AI Smart Frame handles automated editing, and a clipping engine preps broadcasts for release. Manual formatting has largely yielded to automated video editing at Singapore's broadcaster.
Radiodiffusion Télévision Ivoirienne (RTI) — AI in documentary filmmaking reaches African screens via Imhotep Académie, a Cocody school specialising in cultural-heritage tech, teaming with La 3 to resurrect colonial-era Ivorian musicians on screen. Their short film, Quand l'Afrique chantait sous la censure, forms episode two of L'Ère des Indépendances, deploying digitally rendered figures for historical narration to chart Côte d'Ivoire's independence-era story.
du — du has switched on an Arabic telecom LLM built solely for in-house use, co-created with Microsoft, Nokia, Khalifa University and the ITU. Rather than facing customers, the Large Language Model runs on the UAE operator's own systems so its engineers and support teams can triage device faults, field complaints and draw operational insights, all in Arabic. du calls it the first system of its kind regionally, part of a wider move toward telecom natural language processing.
WELT TV — Germany gained its first fully AI-produced news broadcast when WELT TV launched KI-WELT, fronted by an AI news anchor modelled on WELT correspondent Paul Klinzing. This AI-generated news presenter tackles weekly robotics and digital-trend segments, yet human journalists retain final editorial sign-off before anything airs.
KT — Five seconds off every call and 4,500 staff-hours reclaimed monthly: that's KT's payoff from an AI customer service deployment running its own generative AI model, Mi:um, across KT, K Bank and KT M Mobile. Wrap-up paperwork, once a third or more of an agent's shift, now takes barely 3%, KT says.
Paradise Mobile — Bermuda's Paradise Mobile built Skippy, an internal assistant giving staff instant answers on customer service, technical support and sales. Pari, its customer-facing counterpart, answers website visitors, part of the operator's push into agentic AI in telecommunications. Paradise Mobile brands itself among the earliest telecoms worldwide embracing this agentic AI for telecom operations model, an unverified claim.
Grupo OPSA — MarIA (Módulo Autónomo de Revisión con Inteligencia Artificial) handles daily editing, proofreading and SEO at Grupo OPSA's EL HERALDO and La Prensa, a case of AI in journalism in Honduras. Honduran staff built it via WAN-IFRA's LATAM Newsroom AI Catalyst scheme, using each outlet's style rules. WAN-IFRA later ranked the group among Latin America's four most inventive media firms for this newsroom AI push.
Infobae — All 478 reporters at Infobae must now run their work through ScribNews, spanning six newsrooms in Argentina, Mexico, Colombia, Peru, Spain and Miami. This AI journalism tool packs 25 assistants that handle SEO tags, write captions, flag keywords, pull images and produce podcast audio. Leadership issued a top-down directive folding generative AI for news production into daily routine, leaving no option to opt out.
La República — A Colombian business daily let ChatGPT run its entire newsroom for one day, producing what its director calls an AI-generated newspaper edition unlike anything the country had seen before. Fernando Quijano, who leads La República, said the chatbot drafted, revised and ranked all 43 stories, 49 photographs and 15 infographics, plus columns, captions and headlines, while Gemini joined ChatGPT in generating the pictures. He described it as the first fully machine-assembled issue from any Ibero-American paper — a milestone for AI in journalism.
Capital (Economedia AD) — AI voice cloning for articles now powers spoken versions of stories at Capital, Bulgaria's leading business title owned by Economedia AD. It tested several vendors before picking one able to reproduce Bulgarian pronunciation. A new Storytel subscription brings AI-generated audio for news articles to readers, plus themed article playlists.
NOESIS Science Center and Technology Museum — Greece's first AI visitor assistant, NOE515, now stands at the ticket counter of Thessaloniki's NOESIS Science Center and Technology Museum, an AI-powered guide for museum-goers still being trialled in beta form. Built atop Google's Gemini and hosted locally rather than in the cloud, it quizzes each party on numbers attending, hours free and personal interests, pairing their answers with a fitting ticket and suggested walking route. Staff wrote the code in Go and HTML, aiming to switch it on fully come September 2025.
Monaco Telecom — Altores, a Monaco-based startup, teamed with Monaco Telecom to build a multilingual AI virtual assistant for customer service, live on the operator's website and WhatsApp. It fields questions non-stop, 24 hours daily, then hands trickier cases off to a human agent. The rollout gives Monaco's telecoms sector one of its first homegrown deployments of a generative AI customer service assistant.
92 News — Pakistan's first Urdu AI news reporter went live on broadcaster 92 News, presenting bulletins in the national language to a mainstream television audience. Live updates and pre-written segments both come from the digital anchor, the channel says. The "first-ever" claim remains unverified independently.
e& UAE — e& UAE has deployed AI-powered autonomous drones for telecom tower inspections nationwide, monitored from a central operations hub. The drones spot structural faults and feed results into e&'s infrastructure platform for predictive maintenance. This AI-guided drone inspection pairs live analytics with autonomous flight to boost safety.
TV JOJ — Slovakia's TV JOJ runs Adam, a self-built AI avatar hockey presenter, at the 2025 Ice Hockey World Championships. Adam handles fan queries on teams and players, forecasts outcomes, and shows up live on air plus across social posts. TV JOJ next wants this AI-powered digital presenter stretching past hockey.
Ooredoo Qatar — A new virtual assistant called Obot now handles support duties for Ooredoo Qatar, built on GPT-4o and reachable through the Ooredoo App and website, ahead of a planned WhatsApp and social media rollout. This is an example of an AI chatbot in telecom customer service, fielding questions and recommending products while refining replies from ongoing interactions. Ooredoo introduced the system at Web Summit Qatar 2025 as part of its wider AI-driven customer engagement strategy.
VG (Verdens Gang) — Norwegian readers can now put questions straight to heiVG, the AI chatbot news tool Schibsted's VG has switched on for its full audience following a testing phase. Behind the scenes it merges OpenAI's GPT-4 with VG's own stored coverage, so replies stay grounded in what the paper has actually published on sport, culture and current affairs. Schibsted Media AI Labs co-developed the conversational news assistant alongside Aftonbladet, its sister title in Sweden.
France Télévisions — MedIAGen now gives 1,200 journalists at France Télévisions secure access to a private generative AI assistant for journalists, built on the broadcaster's own cloud infrastructure, with all 9,000 staff due to gain entry by October 2025. Staff can upload documents spanning several hundred pages and interrogate them through a RAG setup pulling on OpenAI, Google and Mistral models. On France Info, the French public broadcaster also runs automated voice-to-text subtitling, stretching live captioning 12 hours beyond what manual work alone covers.
Hormuud Telecom — Somalia's biggest telecoms firm, Hormuud Telecom, built an AI assistant in Somali called Caaqil, reachable purely by text message where smartphones or data plans are scarce. Dial *1000#, text a question, and get replies spanning health, farming, schooling, money matters and general topics. Basic handset owners, the bulk of the country's mobile users, are the intended audience for this AI-powered SMS service.
Teleuniverso Canal 29 — A first among Cibao regional broadcasters, Teleuniverso Canal 29 in Santiago now embeds AI in newsroom operations as standard practice, ranking among the earliest nationwide to do so. Automated tools now handle social media content generation, editorial optimisation for the outlet's website, and newscast script writing, marking a shift from occasional use to routine editorial infrastructure powered by artificial intelligence in news production.
Ooredoo Palestine — Moeen (معين), Ooredoo Palestine's AI customer service chatbot, now handles more than 70% of customer questions without staff involvement. Reachable through WhatsApp or the company's site, the virtual assistant for customer queries manages tasks like checking balances, paying bills, activating services and updating accounts, forwarding trickier cases to human staff. Ministry of Telecommunications and Digital Economy officials heard about the rollout at the Palestine Digital Conference.
Circuito Líder 94.9 FM — Circuito Líder 94.9 FM in Caracas became Venezuela's first station to run AI-generated radio content without any human hosts, launching 'Vacílate Esto IA' on weekdays. The format clones the voices and personalities of its creators, systematising output through Perplexity and ChatGPT.
Maxis — Roaming help, destination advice and custom suggestions reach Maxis's B2B accounts 24/7 via Mia (Maxis Intelligent Assistant), a generative AI concierge — an AI-powered virtual assistant on AWS's Amazon Bedrock. Reached through the Maxis Business Hub portal, it handles routine queries without human staff.
News Corp Australia — NewsGPT now runs across News Corp Australia's editorial and business teams, blending OpenAI's ChatGPT, Google's Gemini and Anthropic's Claude into one enterprise generative AI tool for newsrooms. Journalists and staff use this AI-assisted journalism workflow to draft copy, condense lengthy files and speed everyday tasks. Access opened in March 2025, reaching every employee within days.
KDDI — KDDI brings AI in customer service to its call centre with au Support AI Advisor, merging a 3D digital-human avatar and generative AI to read voice, text and image queries. Ponta-point questions launch first; full au service coverage follows by fiscal 2025's end. It joins existing automated customer support: voice recognition since 2023, generative AI chatbots since 2024.
KBS — South Korea's KBS has proclaimed 2025 the launch year for its AI broadcasting push, weaving automated tools into news, animation and emergency alerts. A synthetic-voiced radio host now anchors bulletins on KBS Hanminjok Broadcasting, and a generative AI animation series draws on the classic horror drama 'Legend of the Hometown'. Since 2023 the broadcaster has run a CCTV network of roughly 12,000 cameras nationwide to catch earthquakes and floods automatically, a setup now broadened for AI-powered TV production so wildfires trigger the same automatic alerts. KBS has additionally drafted an ethics guideline steering how artificial intelligence gets used across its content.
Reshet 13 — Kaltura's AI TV Genie and Content Lab now compress Reshet 13's hour-long evening bulletin into brief, focused clips, an instance of AI video production in journalism handling footage indexing and themed reel assembly. The Israeli broadcaster's site draws 3.5 million unique visitors monthly. Reshet 13 joined over 50 organisations testing this automated news-clip generation software at its March 2025 launch.
Umniah — Umniah launched an AI call analysis service with Globitel at Mobile World Congress. This AI-powered contact centre analytics tool scores agent performance and caller sentiment via tone detection, feeding dashboards for service centre monitoring.
Tele2 Latvija — Dzirkstele, an AI customer assistant, now handles shopper queries at Tele2 Latvija's online store, covering internet plans while drafting device descriptions unaided. Sentiment, language and context analysis power its replies. Tele2 calls this virtual assistant just the launch stage of a larger AI rollout.
Clip News — Clip News, a Cyprus media intelligence firm, runs an AI-powered radio and TV monitoring platform across 39 stations, 24/7: 15 in Cyprus, 24 in Greece. Built with Italian partner Digital Wave Form, it applies Automatic Speech Recognition above 90% accuracy for instant alerts and transcripts for marketing and communications staff, marking Greece's first use of automated broadcast transcription monitoring.
Radio Television Hong Kong (RTHK) — RTHK ran CPII's AI media assistant a second straight year for its Budget Speech broadcast, live-transcribing and translating the Financial Secretary's remarks in Cantonese, English and Mandarin. A new chatbot took audience questions while scored summarisation joined the rolling summaries carried over from before.
New York Times — Echo, a new summarisation system from The New York Times, condenses articles, briefings and interactives amid a wider rollout of generative AI tools newsroom-wide. Staff can also draw on GitHub Copilot, Vertex AI, NotebookLM and OpenAI's API for SEO headlines, social posts, edit suggestions and interview questions. This AI in journalism push keeps Echo in beta.
NetOne — Zimbabwe's NetOne embedded an AI chatbot, Tutsirayi, inside WhatsApp, letting subscribers refill airtime and adjust accounts unaided, bypassing agent queues. This self-service assistant channels OneMoney, EcoCash and PayPal payments, easing pressure on call centre staff.
MTN Zambia — MTN Zambia's Call Natasha, built with blackNgreen, automatically handles up to 83% of queries, cutting response times and costs, MTN says. This AI customer service tool answers financial tips, news and general questions in real time, an automated support tool covering everyday account queries in Zambia. The 83% figure is MTN's own, unverified.
Indosat Ooredoo Hutchison — Indosat Ooredoo Hutchison's around 14 AI use cases lifted network prediction accuracy from 70-72% to 88%, saving approximately USD 25 million in six months. The Indonesian AI in telecom work, under the AI North Star strategy, covers churn reduction, network optimisation and customer experience. Indosat targets nearly 200 telecom AI use cases.
Radio Tel Aviv (102FM) — 102FM Tel Aviv gave Odelia AI a daily 2pm-4pm afternoon berth, the first AI radio host anywhere in Israeli broadcasting. Songs, traffic bulletins and comic banter fill its airtime, part of a wider lineup shake-up the station touts as groundbreaking for AI-driven radio presenting.
El País de Cali — 12 months of verified journalism now feed an AI chatbot news archive tool at El País de Cali, named "Pregúntale a El País", which lets readers pose questions and get replies linked back to the original stories. Engineers built this AI-powered news assistant by vectorising the paper's historical reporting and applying embeddings, creating the retrieval foundation behind it. Bosses intend to push the assistant beyond the web onto WhatsApp, and a sponsored-question advertising model is already taking shape.
Magyar Telekom — Hungary's Telekom app now offers an AI assistant telecom app feature called Magenta AI, drawing on Perplexity search technology since May 2025. Live web answers rely on GPT-4.1, Claude Sonnet 4.7 and o4-mini, whereas customer support questions get forwarded onward to the existing Vanda chatbot. Magyar Telekom positions this as everyday public access to an AI-powered search assistant, aimed at the wider consumer market.
Vaduzer Medienhaus — No human touches the microphone at Radio Vaterland, an AI radio station run by Vaduzer Medienhaus. Software condenses newspaper articles into spoken summaries voiced through cloned copies of staff members, while also curating the soundtrack. This automated approach to radio broadcasting removes editorial and presenting jobs from the on-air workflow entirely.
Astro Malaysia — Astro's sooka streaming service runs on ThinkMediaAI, delivering AI personalisation streaming for recommendations, curation, and metadata enrichment. The Malaysian broadcaster credits the rollout with higher usage and viewing time, though the gains are self-reported, not independently audited. AI-powered content recommendation now drives sooka's engagement.
Liberty Puerto Rico — At Liberty Puerto Rico and the US Virgin Islands, an AI customer service telecom platform screens each invoice ahead of dispatch, holding precision at 99.9%. LiBi manages payments and requests, sensors detect internet glitches within two minutes, and chatbots back up phone support teams. Managers link these AI billing accuracy improvements to healthier NPS and churn nearly halved.
AT&T México — Wim gives AT&T México a mobile brand built on AI-powered customer service chat with no human agents. AT&T Ask, its internal generative AI tool, serves finance and strategy staff. Both back AT&T's 'digital telco' push on AWS cloud infrastructure via conversational AI customer service.
Presidential Broadcast Service-Bureau of Broadcast Services (PBS-BBS) — Radyo Pilipinas rolled out AI in broadcast journalism Saturday: virtual hosts Aivan and Aira front its AI Talks show for the Philippine state network. Staff at the Presidential Broadcast Service-Bureau of Broadcast Services (PBS-BBS) vet every line before the AI-generated anchors speak it, said Fernando Amparo Sanga, who pitched the pair as a fix for wider reach against fake news.
Thai PBS — Thai PBS has launched three AI-driven features on its news website, blending synthetic voice, story digests and content recommendation into one push for AI news personalisation. News out Loud reads out articles by mixing machine-generated speech with human anchor tones, while AI in Brief compresses stories into quick summaries. A third tool, Playlist for You, tailors suggestions to individual viewers through VIPA, the broadcaster's OTT streaming platform, delivering personalised recommendation and summarisation. Kuala Lumpur's ABU Con-Fest served as the stage where Thai PBS unveiled the trio, tied to a broader accessibility drive in digital journalism.
Mobiuz — Uzbekistan's Mobiuz fields 25 to 30 per cent of call-centre traffic via its AI voice assistant for customer calls, Moobina, resolving tens of thousands of daily calls. Atop the Omilia AI platform, this AI customer-service bot confirms caller identity, verifies balances, guides tariffs and unblocks numbers unaided, work begun in 2024.
Ålandstidningen — Ålandstidningen, a Swedish-language paper in the Åland Islands, offers live AI sports streaming of football, ice hockey and volleyball. One staff member positions a three-lens camera at each venue, and the AI-driven broadcast captures audio and pictures unaided.
KLM — KLM's AI email marketing tool Airmail has sent 50 newsletters to 6.3 million customers, cutting email creation time by 90% with no drop in customer satisfaction scores. Airmail lets the Netherlands airline's local marketing teams assemble AI-generated campaigns from nearly 60 destinations and five themes.
Asahi Shimbun — Asahi Shimbun deployed Microsoft 365 Copilot across nearly all departments after a four-month pilot with 300 staff found roughly two hours saved weekly per person. This AI in journalism rollout followed an AI Committee formed that April by its CEO, backed by a company-wide "AI full-commitment" mandate; adoption reached over 90% of staff within five months. Agent-building tools within Microsoft Copilot Studio are now under trial, extending generative AI in newsrooms into daily reporting.
SoftBank — SoftBank has deployed voice AI customer support for Y!mobile PIN code inquiry calls in Japan, using an autonomous-thinking generative AI platform. Reading speech from several angles for context and intent, the platform questions callers and verifies details; a large language model plus a proprietary knowledge-reasoning engine handles complex conditional judgments and non-standard inquiries. SoftBank has used call-centre AI since 2019, co-developing generative AI platforms with Microsoft since March 2024.
Mobicom Corporation — AI-based customer personalisation has arrived in Mongolia's telecom sector via Explore, built by Mobicom Corporation with partner Circles. Usage patterns feed one-to-one promotions to curb subscriber loss across a market of 3.5 million people. Smaller operators increasingly use such AI-driven personalisation tools to cut churn.
Topnet — Topnet, a Tunisian internet service provider, now runs Topy, offering AI customer service in telecommunications 24 hours a day, 7 days a week through an AI virtual assistant for customer support. It gives instant, tailored replies on billing and technical issues, plus visibility on network status. Topnet says the goal is faster responses, with the assistant working alongside, not replacing, human support staff.
Reach plc, the UK's largest commercial news publisher, uses Guten, an in-house AI tool that turns trusted news-wire copy into stories for its titles, which journalists review before publication. At WAN-IFRA's Newsroom Summit, chief digital publisher David Higgerson said Guten cut breaking-news publishing from nine minutes to 90 seconds and supported about a quarter of output.
Digicel — Digicel's AI customer care assistant Ruby has handled 55% of more than 535,000 conversations in Jamaica without a human agent, the telecoms group's 2026 Sustainability Report says. Built on Canadian firm Ada's platform, the virtual assistant takes routine jobs such as bill payment, plan activation and balance checks across voice, web, WhatsApp and the MyDigicel app, supporting care teams without replacing them, says Digicel.
Only 13% of 280 news leaders in 51 countries called their newsrooms' current AI initiatives 'transformational' in the Reuters Institute's 2026 survey (44% promising, 42% limited). The report flags growing 'disillusionment', a sober turn after last year's claim that most newsrooms were transforming.
85% of researchers in a Wiley survey of 2,430 (August 2025) credit AI with better efficiency. Nearly three-quarters saw gains in both quantity and quality of work.
University of Aden — Dr Sahar Muhammad Ahmad Ali, assistant professor at the Faculty of Media, University of Aden, puts AI in investigative journalism at 57% among Yemeni practitioners, from a digital questionnaire answered by 30 reporters. Dependence rates as middling for 53%, with hunting down information and checking facts the top applications and ChatGPT the favoured tool. Barriers to AI-assisted reporting: poor instruction, at 93%, and frail digital infrastructure, at 86%.
NOVA FCSH / Calouste Gulbenkian Foundation — NOVA FCSH's White Paper on Artificial Intelligence in Journalism finds 83.8% of journalists nationally have never been trained in AI, and 64% work in newsrooms lacking an AI code of conduct. Funded by the Calouste Gulbenkian Foundation and presented in Lisbon, Portugal's first national study of AI in journalism sets out 10 strategic recommendations for ethical newsroom AI adoption.
Quatromanos — Quatromanos polled over a hundred Uruguayan reporters: 60% have relied on AI in journalism, 26% habitually and 37.5% occasionally. Leading uses were converting interview audio to text (41.7%), background research (20.8%) and brainstorming angles (13.5%). Most AI-assisted reporting sits in groundwork, not the writing.
Centrul pentru Jurnalism Independent — Two-thirds of Moldovan newsrooms, 67 per cent, now use AI in journalism, a Centre for Independent Journalism poll of 101 outlets shows, reflecting broad newsroom AI adoption. Leading uses: automatic translation, 63 per cent; text generation, 57 per cent; automated proofreading, 54 per cent.
Suomen Lehdistö (Finnish Media Federation) — AI in newsrooms now touches nearly every desk across Finnish media, a survey revealed. Viestimedia's Renki raises editorial productivity by 20% through 25 built-in capabilities, and Alma Media runs close to 20 separate systems spanning headline writing and transcription. Roughly half of Yle's reporters lean on YleGPT weekly as their AI aide, whilst Bauer Media pipes radio bulletins through automated voices across 11 channels — AI adoption in media houses reshaping daily production.
Fiji media sector — Over 50 per cent of surveyed media workers in Fiji already use AI in journalism, relying on tools such as ChatGPT on the job, per ABC International Development's State of the Media: Fiji 2025 report. Editing, research and drafting social posts topped their listed uses, though many flagged worry about losing a human touch, showing cautious AI tools adoption among newsroom staff.
Researchers' AI tool use rose to 84% in 2025 from 57% in 2024. Survey: 2,430 researchers, August 2025.
Malta Gaming Authority — Malta AI regulation gained a gambling rulebook as the Malta Gaming Authority and the Malta Digital Innovation Authority unveiled the voluntary AI Gaming Charter. Licensees receive principles for accountable, transparent machine learning, shaped by industry polling, licensee talks and formal consultation. Operators already use such systems to protect players, spot fraud, serve customers and guide decisions; these AI rules for gaming sit beside the EU AI Act.
South Sudan Media Authority — AI in newsrooms now has official backing in South Sudan, as the Media Authority tells outlets to embrace the technology carefully. Speaking in Juba, Sapana Abuyi said nothing bars its use, but pushed for training staff on verifying AI-generated material before publishing. News came via a gathering run by the Union of Journalists of South Sudan and Norwegian People's Aid.
Congreso de la Unión / Diario Oficial de la Federación — Mexico's Federal Copyright Law reform, in the Official Gazette on 14 May 2026 and effective next day, imposes an AI voice cloning ban on performers' voices and faces without consent. Fines reaching 40,000 days' minimum wage shield voice actors, dubbing artists and announcers from unauthorised clones.
National Assembly of the Republic of Armenia — Fines of AMD 300,000 now apply under Armenia's RA Law on Audiovisual Media, amended to force AI content labelling. Providers must flag generative AI or machine-learning output as 'Created by AI', or 'Digitally processed' if merely altered. This synthetic content labelling duty catches subtle edits, not just wholly generated media.
Conseil de Presse (Tunisian Press Council) — Tunisia's Press Council has drawn formal boundaries for AI in journalism, launching the National Charter for AI Use in Journalism on 15 March 2026 to sort applications into allowed, borderline and banned categories. Signatory newsrooms must disclose AI-generated material to readers and keep a human reviewing every automated output. Compliant outlets earn a trust label for artificial intelligence in news production.
Ministry of Media and Saudi Data and Artificial Intelligence Authority (SDAIA) — Saudi Arabia's Ministry of Media and SDAIA launched the AI Principles in Media document at the fifth Saudi Media Forum on 3 February 2026, establishing formal guardrails for AI in media nationwide. Eight principles anchor the rules governing content creation, editing, publishing and distribution, spanning disclosure of AI-generated material, privacy safeguards, credibility, information integrity and a ban on misleading content. This marks artificial intelligence deployment in journalism now under structured national oversight.
Haute autorité de l'audiovisuel et de la communication — The Haute Autorité de l'Audiovisuel et de la Communication forbade newsrooms to doctor what goes to air, or what candidates say, using “uncontrolled and irresponsible” artificial intelligence. That Benin AI regulation shaped the presidential ballot of 12 April 2026, the opening round's canvassing window running 27 March to 10 April. Outlets straying from the AI campaign rules risk penalties reaching instant removal from official election coverage.
Conseil National du Patronat de la Presse et de l'Audiovisuel (CNPA) — Benin's media sector now follows a 33-article Code of Ethics covering AI in journalism, swapping out rules that had stood for 25 years. Reporters must check facts before publishing and resist disinformation, while new guidance points artificial intelligence in newsrooms toward safeguarding integrity and ethical conduct. A countrywide drafting effort, spanning the diaspora, shaped the update.
Utilities Regulation and Competition Authority (URCA) — Bahamas broadcasting regulation now compels licensed operators to draft or update governance rules for artificial intelligence, synthetic media and deepfake tools. The Utilities Regulation and Competition Authority delivered this shift through a revised Code of Practice for Content Regulation (ECS 78/2024, 20 December 2024), weaving AI-generated content rules into broader service standards rather than issuing a standalone deepfake measure.
Ministry of Media and Saudi Data and Artificial Intelligence Authority (SDAIA) — Saudi Arabia's Ministry of Media and SDAIA opened the Artificial Intelligence Center for Media (AICOM), the Middle East's first such body, launched by the Minister of Media and the SDAIA President at the Saudi Media Forum. It tracks AI in media, forecasts coming applications of artificial intelligence in media for journalism, and backs the Kingdom's bid to lead the sector.
Integrated Transport Centre (Abu Dhabi Mobility) — Abu Dhabi's Integrated Transport Centre unveiled Dune at LIVEX 2026, an AI transport planning platform built with Space42 around the STEAM+ model. Planners test mobility policies, infrastructure investments and public transport projects.
Northeast Frontier Railway — Northeast Frontier Railway (NFR) launched Project DRISHTI, AI freight wagon inspection built with IIT Guwahati, at Maligaon on 25 September. Checkpoint cameras read wagon numbers and check doors, locks and seals for tampering or theft. Hand checks, the General Manager said, struggled at night and in bad weather, which the computer-vision screening addresses.
Seara (JBS) — Seara, part of JBS, has put the uMov.me AI delivery assistant Mari on routes worked by 300 drivers, 2,000 vehicles and 20 distribution centres that handle about 500,000 drops monthly. Drivers speak to Mari through WhatsApp voice notes; the tool tracks each truck using existing company data and alerts sales and back-office teams. Seara wants this route-tracking AI to push successful deliveries to 98.5%.
US Federal Aviation Administration — The Federal Aviation Administration began limited operations of SMART, an AI air traffic tool, on Monday 21 September at Ronald Reagan Washington National, Washington Dulles International and Baltimore/Washington International Thurgood Marshall airports. SMART (Strategic Management of Airspace, Routes, and Trajectories) fuses 200 feeds, weather and staffing included, and proposes schedule tweaks. FAA staff vet each one; facility managers decide whether the AI scheduling aid's advice is used.
Autostrada del Brennero — Autostrada del Brennero is trialling AI traffic management for 12 months on 87 km of Italy's A22, Brenner Pass to Bolzano South, via Valerann's Lanternn. CCTV, sensor, weather, floating-vehicle and historical-incident feeds form one live control-room picture. This motorway AI pilot should flag incidents automatically, rate risk and quicken operator responses.
Loblaw Companies Limited — Loblaw Companies Limited is widening AI gate automation to more yards and distribution centres, deepening a tie-up with Canadian firm EAIGLE. EAIGLE's Vision AI, branded AVAC, reads vehicles and captures freight data as trucks arrive, feeding warehouse, transport and yard management software plus ERP. Earlier sites trimmed processing times, smoothed the driver experience and sharpened data accuracy, which prompted the wider automated gate rollout.
Ministry of the Interior and Safety — South Korea's Ministry of the Interior and Safety has settled a Chuseok Holiday Safety Management Plan pushing AI road safety warnings across 50 crash-prone stretches chosen by AI big data. Those accident-risk alerts surface on the T map navigation service and on roadside electronic signboards through this year's four-day break.
CEVA Logistics — CEVA Logistics is trialling driverless pallet movers at Blue Hub: two Zelostech Z10 autonomous electric vehicles on existing multi-level ramps, shifting stock between floors. CEVA expects at least 4,600 kilograms a year in cut emissions. Zelostech met CEVA via the 2025 CMA CGM Startup Awards; this warehouse automation trial decides whether the vehicles reach other CEVA sites.
Hanjin — Major safety breaches have fallen roughly 60% at Hanjin Logistics sites running AI Safety Guardian, an AI camera system for warehouse safety now active across about 20 logistics terminals. CCTV feeds are scanned for staff crossing conveyor belts, vehicles flouting site rules, cars blocking pedestrian walkways and unworn hard hats; violations trigger flashing lights, voice warnings and a text alert to safety managers, delivering surveillance-based hazard alerts in real time. Ten locations near greater Seoul launched first, in late July, before expansion into Chungcheong, Gangwon and other outlying regions.
Verne — Verne and Pony.ai put passengers into robotaxis with an empty driver's seat on Zagreb's public roads from 10 September 2026, which both firms call a European first. Driverless robotaxi rides trace a 22-kilometre corridor from Verne's base and a main city business district to Franjo Tudman Airport Zagreb. Since paid operations opened in April the autonomous fleet has covered over 200,000 kilometres and several thousand fare-paying rides, at a mean passenger score of 4.7 out of 5.
Brisbane City Council — Brisbane City Council said it will hand traffic signal timing at 37 intersections to AI — 22 along Logan Road and 15 along Kelvin Grove Road — in a six-month trial beginning early in 2027. The AI traffic signal system is to read traffic volumes and queue lengths at each intersection and work out the best light sequence from them, learning over time, suggesting interventions to operators and adapting to weather and major events. The council's smart traffic signals plan follows a separately announced trial at one intersection in Petrie, north of the city.
IJM Corporation — IJM opened an RM18 million Smart Highway Traffic Control Centre at the Besraya Loke Yew toll plaza lay-by, watching its Besraya, New Pantai Expressway and Kajang-Seremban concessions through 144 CCTVs, 87 video incident detection systems, 35 automatic number plate recognition scanners and 22 variable message signs. The AI highway monitoring flags congestion, pedestrians on the carriageway, early flood signs, liquid spillage, potholes, dumped rubbish, wrong-way and reversing vehicles, illegal lane changes, hard-shoulder stops and illegal advertising banners and posters, and counts vehicles by class. Works Minister Datuk Seri Alexander Nanta Linggi officiated, and the centre's smart highway data feeds the Malaysian Highway Authority's national traffic management centre.
Uganda Airlines — Amadeus will supply Uganda Airlines with SkyWORKS, an AI scheduling in aviation platform, making the carrier the first in Africa to adopt the system. It brings schedule editing, feasibility checks and AI-driven flight planning to support the airline's growing regional and long-haul routes.
Ethiopian Shipping and Logistics — Ethiopian Shipping and Logistics unveiled Digital ESL, an AI logistics platform for web and mobile, in Addis Ababa. Importers, exporters and partners place orders, schedule container loading and follow cargo any time, anywhere; the state-owned operator says AI algorithms remove human handoffs. The self-service system replaces paper clearance at Ethiopia's dry ports and sea supply routes.
CJ Logistics — CJ Logistics has deployed a pair of dual-arm humanoid robots for live packing at the Olive Young logistics centre in Yangji, Yongin, marking its debut for humanoid robots in logistics. The dual-arm machines slot cushioning into boxes as orders are packed, and the company plans to extend them to inspection, sorting and picking while feeding site data back into training their control models.
Uber — Uber and Wayve have launched supervised autonomous rides in London, the first self-driving trips offered in the United Kingdom. Riders booking UberX, Uber Electric or Uber Comfort may be matched at no extra cost with a Wayve car, an all-electric Ford Mustang Mach-E with Wayve AI Driver and surround sensors. A trained, TfL-licensed private hire driver oversees each trip, and a small fleet serves the city at first, airports excluded.
Djibouti Ports and Free Zones Authority — Djibouti Ports and Free Zones Authority has revealed DPCS Foresight, an AI port logistics tool that reads supply-chain records, history and operational performance to forecast when importers will fetch containers. Yards park imminent departures nearest the gate, burying slower boxes, so container pick-up prediction trims re-handling and dwell times. The reveal came at Djibouti's debut national artificial intelligence forum.
NJ Transit — NJ Transit's redesigned website now carries Navvie, an AI chatbot giving riders conversational answers without navigating multiple pages, the agency says. Navvie answers questions on schedules, alerts, transfers and trip planning, and may be added to the NJ Transit mobile app by end-2026. It answers Governor Mikie Sherrill's executive order on improving digital infrastructure and passenger information; a separate request for information seeks a unified real-time customer information platform.
City of Moreton Bay — The City of Moreton Bay will run Australia's first AI traffic lights trial at the junction of Paper Avenue and Moreton Parade in Petrie, using SWARCO software that controls signal movements in real time on active demand instead of stepping through fixed phases. Cameras and 3D LiDAR sensors are to tell passenger vehicles, heavy transport, cyclists and pedestrians apart, with the system learning local patterns and recalculating sequences on the spot. Testing was set to start in September 2026 and the AI signal control trial runs to 2029; no source confirms the switch-on day.
Eastern Pacific Shipping — Eastern Pacific Shipping is extending AI collision avoidance from Orca AI to dozens more ships after a five-vessel pilot of over 130,000 nautical miles. The companies report 48% fewer close encounters per 1,000 nautical miles and 10% wider average passing distances, figures with no independent evaluation. Its computer-vision SeaPod bridge unit picks out and classifies surrounding objects and vessels, even craft without AIS.
ClassNK — ClassNK has launched 'Survey Compass', a natural-language AI ship survey assistant answering shipowners' and managers' classification-survey questions via its customer portal. For accuracy in safety-critical work it relies on retrieval-based answers drawn from, and linking back to, source documents instead of generating unsourced text.
Tocumen International Airport — Tocumen International Airport's rebuilt digital platform features Tocumín, an AI airport assistant that tells a traveller at gate 115 how long the walk to gate 205 takes and the route, alongside retail, parking, connection and flight details. Tocumen S.A.'s in-house team built the wayfinding assistant in about 18 months; biometric access to travel information is due in 2027. The airport handled 13.7 million passengers from January to July 2026, 15 per cent up on a year earlier.
Air Transat — Air Transat, launch airline for Montreal firm stratosX, runs its X360 decision-support platform in its operations centre. Combining predictive modelling and generative AI with the airline's operations data, it recommends aircraft and passenger rerouting through schedule changes, delays and cancellations, with crew, aircraft and passenger data in one view. Air Transat co-develops it with stratosX and expects faster recovery, clearer cross-team visibility and steadier decisions.
Sharjah Roads and Transport Authority — The Sharjah Roads and Transport Authority is fitting AI bus cameras to 131 intercity buses, ten each with six inside and four outside, to catch driver fatigue or inattention, count boarding and alighting passengers and log accidents and unusual incidents. Taxis already carry five AI-supported smart cameras watching the road and the cabin. Chairman Yousef Khamis Al Othmani says the rollout continues for safety, efficiency and customer experience.
Delta Air Lines — Delta Air Lines chief executive Ed Bastian told a podcast the airline will expand AI in ticket pricing and predictive operations, expecting up to 50% better profitability. Delta calls its pricing system a decision-support tool on aggregated market data and denies individualised fares based on personal data. US lawmakers want answers on how AI and personal data affect fares.
Aeropuerto Internacional de Punta Cana — Punta Cana International Airport has deployed SeeTrue's AI baggage screening, which reads Leidos ClearScan CT scanner images at checkpoints live to help inspectors find possible illicit substances and undeclared cash in luggage. The airport calls it the first international-airport use of SeeTrue's algorithms with ClearScan for drug and currency detection, built to avoid slowing passenger processing.
Malta Public Transport — San Lawrenz, Gozo hosts Malta Public Transport's autonomous bus pilot, launched 19 August 2026. A safety operator supervises the route between Ċangura and Dbieġi while onboard AI, sensors, cameras, GPS and LiDAR handle the driving. Free rides come with a survey request, helping the University of Malta evaluate the driverless shuttle scheme.
Singapore Airlines — Singapore Airlines has more than 160 airline AI applications in production, among them a Flight Recommender fielding nearly 10,000 queries a day by the end of FY2025/26. 'JARVIS', its generative-AI staff assistant, serves about 5,079 of 5,600 ground staff, per the annual report and senior vice-president of IT George Wang.
Ryanair — Ryanair has signed a five-year deal putting Google Workspace and Gemini Enterprise in front of 35,000 employees, aiming AI in aviation at crew logistics, decision automation and corporate productivity. It also plans Google DeepMind's AlphaEvolve and WeatherNext models for maintenance scheduling and fleet operations; no deployment timeline was given.
PostNL — PostNL is deploying AI agents to automate administrative and support work under what it calls an 'AI-first' strategy, speeding process automation and standardisation. The work sits in a cost-savings plan now up to €180 million, including about €75 million extra across 2027 and 2028. Chief executive Pim Berendsen said this makes the cost base more competitive; no AI products were named.
Sunrise Airways — Haitian carrier Sunrise Airways has put a multilingual AI customer service assistant on WhatsApp to help passengers before, during and after flights. The airline expects the WhatsApp helper to smooth contact with travellers during delays, cancellations and other disruption.
Port Authority of New York and New Jersey — John F. Kennedy, Newark Liberty, LaGuardia and New York Stewart International now share a single AI travel assistant, built into revamped Port Authority of New York and New Jersey pages. It answers queue length at security or customs, free parking spots and nearby amenities, sparing travellers from separate listings. Officials frame the airport chatbot as easing journeys through these United States hubs.
Netzence Sustainability Limited — Netzence Sustainability Limited debuts KlimateRide in Lagos, Nigeria's first AI-driven ride-hailing platform. Onboard software runs allocation, routing and live tracking, plus ID checks, an SOS button and route analysis under this AI ride-hailing and delivery service. Drivers keep 92 percent of each fare.
Roads and Transport Authority (Dubai) — Dubai's Roads and Transport Authority (RTA) says its new AI bus dispatch Decision Support System will cut bus deployment to under a minute from 30 to 60 minutes. Machine learning and optimisation algorithms choose and dispatch the best-suited buses live, watching over 1,100 buses across 26 operational scenarios to match Metro and Tram disruptions.
Nestlé Japan Ltd. — Hacobu's MOVO Vista and MOVO Driver apps form an AI logistics platform now linking roughly 7,000 contracted truck drivers to Nestlé Japan's delivery network as a freight dispatch system. Dispatchers see wait times, arrival and departure schedules, and later location sharing, supporting compliance with Japan's revised logistics efficiency law.
Serbian Railways Infrastructure — Serbia opened a EUR 127 million Unified Dispatch Center at Belgrade's Makiš yard, applying AI rail dispatching to 540 km of operational track, staffed by 23. Russian Railways subsidiary Institut NIAS co-delivered the hub, whose algorithms supplant manual protocols. Centralised, automated train dispatching is planned for all 3,357 km nationwide.
Ministry of Transport of Tajikistan — Tajikistan's Ministry of Transport is piloting an AI call centre for carriers, run by its Center for Digitalization of the Transport Sector and reachable on 8808. It aims to cut the time needed to secure permits, passes, vouchers and cargo waybills. An AI-powered call centre for freight operators like this also comes with a Telegram channel, Centre for Information Support of Users of Transport Systems, and a chatbot, MintransBot, for user queries.
Vilniaus viešasis transportas — Vilniaus viešasis transportas, public transport operator for Lithuania's capital, now uses AI customer service for its 50 to 100 daily passenger enquiries. Built with IT firm Nortus on a Microsoft platform in four weeks, it uses Gemini to read free-text letters, extract key facts, log them in the central database and draft replies.
Dangote Cement — Crashes involving Dangote Cement lorries dropped 56 percent from 2025 to 2026, by the firm's own count, following AI fleet safety tools plus upgraded GPS and telematics in Nigeria. Drivers are watched live and automated inspection technologies check vehicles, an AI-led haulage safety push aimed at efficiency.
Planzer Paket — Mathilde, a Level 4 autonomous delivery vehicle from logistics firm Planzer Paket and startup LOXO, drives Bern's public roads driverless. Electric-powered, it links Planzer's rail hub with city transfer points. An attendant still sits in the passenger seat under a Federal Roads Office permit, with fully driverless delivery the eventual goal.
Ministry of Transport, Communications and Information Technology (Oman) — Oman's Ministry of Transport, Communications and Information Technology has introduced the country's first AI asphalt paving technology on phase four of the Sultan Said bin Taimur Road dualisation in Maqshan, Dhofar. The goal is faster construction at high quality; contractor Galfar Engineering & Contracting says the self-paving system coordinates paving machinery and cuts manual work on site.
Kenya Airways — Kenya Airways has taken on Jupiter 5.0, FlyNava Technologies' system for AI fare pricing. With the AI-driven pricing tool, the airline's revenue team models new fare structures before rollout and responds quicker to shifting demand while pricing calls stay visible.
Western Cape Department of Infrastructure — The Western Cape Department of Infrastructure watches about 5,000 km of strategic provincial roads with Bentley Systems' Blyncsy, the technology's first African use, per Bentley. The AI road inspection tool runs machine learning over crowdsourced dash-camera footage to flag damaged guardrails, missing signs, broken streetlights, debris and encroaching vegetation. Maintenance crews now go where most needed, says the chief engineer for the province's transport infrastructure systems.
Traffic Safety Council of Zimbabwe — The Traffic Safety Council of Zimbabwe (TSCZ) has begun AI vehicle monitoring that captures registration marks instantly, matching each against several record sets and flagging offending cars. A high-definition lens feeds AI software able to fine motorists immediately for unpaid licensing fees. Trialled during April 2026, the plate-scanning road surveillance featured at Mucheke River Bridge's Masvingo opening.
Talabat Egypt (Talabat Mart) — Talabat Egypt runs the Middle East's largest express commerce distribution centre, where in-house AI demand forecasting sets stock and automation trims losses. The 27,000 sqm Yanmu East Logistics Park handles one million items daily, AI-based forecasting feeding live inventory, retail and supply-chain links. Serving Talabat Mart across 12 cities, it houses 74% of shared services staff for eight markets.
Yorozu Logistics Co., Ltd. — Osaka's Yorozu Logistics is trialling AI truck dispatch software with Super Academy, aiming to supplant decisions long resting on veteran schedulers' instinct. It weighs distance, fuel, tolls and driver hours to surface the single most profitable route, a generative AI logistics planning pilot run under live conditions.
Verne — Verne, part of the Rimac Group, opened paid robotaxi rides in central Zagreb on 8 April 2026, billed by its partners as the earliest anywhere in Europe. The battery cars run Pony.ai's seventh-generation autonomous driving system, hailed through the Verne app, with Uber app booking promised under a three-way tie-up between Verne, Pony.ai and Uber; trained autonomous vehicle operators sit aboard while the deployment is young, and coverage is meant to widen beyond the initial central districts. Croatia's autonomous taxi fleet is, in chief executive Marko Pejković's words, proof that 'For the first time in Europe there is a real commercial robotaxi service.'
Mitsui O.S.K. Lines (MOL) — Inspectors have cleared four Stage 2 ships under Japan's MEGURI2040 scheme for commercial sailing as certified autonomous ships Japan now recognises. A new coastal containership joined a ferry, a Ro-Ro carrier and a refitted older vessel, each self-navigating vessel running at Level 4-equivalent standard while shore teams at Fleet Operation Centres oversee them remotely.
Zalando — Zalando is deploying up to 50 AI picking robots from Nomagic across its European fulfilment network, with first units already running in Lahr and Mönchengladbach, Germany, and Verona, Italy. Nicknamed 'Richard', they pick and scan individual items, shoeboxes included, and feed automated pocket sorters, and Zalando, announcing from Berlin, plans more for Rotterdam, Stockholm, Paris and its new Giessen site.
Pos Malaysia Berhad — Zelos, Malaysia's first autonomous logistics vehicle, now runs for Pos Malaysia with Autonomous Logistic Solutions Sdn Bhd. The uncrewed, electric unit suits repetitive, high-volume point-to-point freight. A six-month pilot follows across the postal operator's network, using 5G from Digital National Berhad to validate the driverless vehicle.
Cornelder de Moçambique — Cornelder de Moçambique, which runs the Port of Beira, has deployed Terminal Flow, a home-grown AI container tracking tool for boxed cargo. The cargo-location system shows long-reach forklift operators exactly where each box sits and where its load is bound, says managing director Jan de Vries.
Danish Meteorological Institute — Machine learning now drives ASIP (Automatic Sea Ice Products), DMI's (Danish Meteorological Institute) tool for AI-based sea ice charting across Greenlandic waters, built from satellite imagery. Royal Arctic Line, DanPilot Greenland and Arctic Command crews compared the automated output against hand-drawn charts through the season and reported quicker, more frequent refreshes than the older method allows. Accuracy still slips along coastlines and once melting begins, DMI notes, a gap explored at the 26 March 2026 review of the AI sea ice mapping trial.
Government of Mauritania — Mauritania now runs an AI road safety system nationwide, catching overloaded lorries and missing seat belts on intercity highways as drivers pass. Engineers from the State Digital Agency and the General Directorate of National Security built the cameras and vision software for the Ministry of Digital Transformation and Administrative Modernization, powering this AI-based traffic violation detection. Each breach triggers an automatic alert and lands in a traceable record, sparing officers from checking vehicles one by one.
Grupo CISA — Grupo CISA now equips Metrobús supervisors with AI chatbots for transport supervisors, offering instant guidance on operating rules, insurer liability standards and internal transport protocols whenever incidents arise. Built by a recent IPN artificial intelligence engineering graduate, the software speeds sanction checks and incident fixes for staff running Mexico City's bus rapid transit operations. A fourth tool under construction will scan photographs to detect vehicle damage.
Prosegur Cash — AI cash logistics cut ATM paperwork 70% for Prosegur Cash in Honduras, Guatemala, El Salvador, via optical character recognition, Big Data and AI models, erasing transcription errors. Francisco Galeano, Honduras and Nicaragua, reported operational metrics up 17%, security metrics up 11%.
National Logistics Corporation — Pakistan's National Logistics Corporation now runs AI cargo scanning at two crossings: Sost Dry Port facing China and Gabd Border Terminal on the Iran frontier. Vehicles clear in 22 seconds as the system reads cargo codes and checks structural details, cutting delays and manual customs work via this AI-based border inspection.
Port of Hanko — A sensor-equipped robotic helicopter now patrols Finland's Port of Hanko unaided, recharging at its own dock between missions. RumbleTools built the Skynest AI drone monitoring system for autonomous aerial checks of harbour docks and equipment, removing the need for a human pilot.
Russian Railways — Chelyabinsk-Glavny station now runs an AI shunting routing system, the first of its kind deployed by Russian Railways, as part of its automated rail traffic management push. An onboard module works out the best path for every train or shunting locomotive, weighing current and planned network load while tracking rolling stock live to stop routes clashing. Once set, the digital order passes straight to the operations management system controlling signals, leaving staff to supervise rather than intervene directly. Track and locomotive use tighten up, planning gets quicker, and pulling manual judgement out of routine moves cuts safety risk.
Nova Poshta — Every Nova Poshta outlet now gets a daily photo review under a new AI branch inspection system, examining whether staff uniforms, marketing-material placement and bin levels meet internal rules. R&D Director Yevgeniy Khomenko credited the AI-based compliance monitoring with lowering costs previously spent on senior managers travelling to sites in person, alongside gains in overall service quality.
Norfolk Southern — Norfolk Southern's AI wheel integrity system uses six cameras plus AI algorithms to catch hidden wheel cracks at up to 70 mph. Live at Burns Harbor, Indiana, this wheel-crack detection technology already flagged a cracked wheel and a casting defect serious enough to trigger an industry recall — the sector's first such deployment, Norfolk Southern says.
Island Oil Holdings — Island Oil Agent, a new AI chatbot on Microsoft Teams, now gives Island Oil Holdings staff instant access to manuals and procedures, marking a wider AI deployment in energy operations at the Limassol firm. Selected staff also use Microsoft Copilot and ChatGPT Plus to draft documents. Robotic process automation bots handle bank reconciliations, and AI-powered cybersecurity tools detect threats and trigger automated remediation, an approach mirroring broader AI adoption in oil sector firms.
City of Tampere — Finland's first commercial automated bus route opened on Nysse line 301 in Tampere, built by the City of Tampere, REMOTED and Tampere University. An OHMIO LIFT vehicle with LiDAR and cameras runs daily under a safety driver, moving toward fully driverless shuttle service at Level 4 by spring 2026.
Correos Express — Correos Express deployed an AI digital twin, built on Neurored TMS & SCM, to sharpen transport network optimisation between Spain and Portugal. Real-time monitoring now spans 70 depots and 550 daily feeder routes, catching congestion points, delays and capacity limits from historical patterns. A future phase will add demand forecasting and automated route optimisation suggestions.
Slovenská pošta — Slovenská pošta is trialling mAIka, its first autonomous parcel delivery vehicle, in central Bratislava, Slovakia. The AI-powered robot serves the embassies of France and Slovenia plus Bratislava city hall. Backing the self-driving postal robot pilot are the Slovak Academy of Sciences, the Ministry of Transport, Bratislava's city government and a technology company.
Sürat Lojistik — AI route planning logistics now spans Sürat Lojistik's network in Türkiye, linking smart lockers to intermodal transport for full shipment tracking. Live traffic and weather feeds drive algorithmic decisions on fuel, demand and routing, while a shared platform unites warehousing and distribution under AI-powered route optimisation.
Southeastern — South Eastern Railway has widened its use of AI-powered thermal camera monitoring, adding the sensors to 20 further Class 375 trains beyond the 12 Class 707 units already carrying them. One Big Circle built the onboard system for the Southeastern and Network Rail venture, spotting overheating rail and infrastructure defects before they worsen; trials so far have caught 41 such faults. That early thermal imaging fault detection on trains has reportedly spared travellers more than 9,000 minutes of lost time.
Aeroporti di Roma — Rome's Fiumicino airport now runs an AI virtual assistant for airport passengers, engineered by Aeroporti di Roma with AWS and Storm Reply atop a multi-agent Amazon Bedrock base. It fields flight, luggage and parking questions plus tailored tips, reachable via WhatsApp or the ADR site in various tongues — an airport chatbot marking ADR's shift toward automated passenger care.
FairPrice Group — Singapore's Land Transport Authority has cleared FairPrice Group as the first local operator permitted to run autonomous delivery vehicles on public roads for freight. Zelos Technology supplies the driverless Zelos Z10 units, which will move stock among FairPrice's warehouses under this approval. FairPrice intends to grow the fleet toward 30 such vehicles, close to 30 per cent of its total transport fleet, in this driverless goods transport rollout.
De Lijn — Ghent-based firm Otiv engineered the tool now running on Antwerp trams for De Lijn, aiming to cut collisions by 20 per cent. Cameras and sensors flag hazards and alert drivers, an early rollout of AI driver assistance at the Flemish public transport company, working much like a digital co-pilot. Drivers involved wear biometric monitors so researchers can track shifts in alertness and workload.
China Logistics — China Logistics has deployed Liuyun, a 2780-billion-parameter model built with China Telecom, China Mobile, iFlytek and Huawei, in an AI deployment in logistics spanning 40-plus scenarios covering transport, warehousing, supply chains and rail equipment. Company figures credit its multimodal transport engine with lifting customer adoption 9%, boosting order matching 10% and trimming average shipping costs by 5%. National generative AI service registration was granted 2 July 2025, ahead of the model's public unveiling. An AI-powered logistics platform of this scale now sits among China's largest state-backed applications.
ADNOC Logistics and Services — ADNOC Logistics and Services cut vessel service sourcing from three hours to 45 seconds with the GCC's first AI-powered Smart Port Solution, bringing AI in port operations to UAE petroleum ports in Das, Zirku, Mubaraz, Ruwais and Jebel Dhana. Built by Innovez One, the port AI trims turnaround time by up to 90% and should save 3,000 hours yearly and $950,000 in operational savings by 2028.
Mowasalat (Karwa) — Robotaxi deployment Qatar has begun: Pony.ai and Mowasalat (Karwa) now run driverless cars on Doha's streets. Onboard operators supervise as engineers tune the fleet for local heat, roads and traffic. Qatar's Ministry of Transport oversees the trial, tied to Qatar National Vision 2030.
American Airlines — American Airlines' AI-driven flight connection tool holds flights at Phoenix Sky Harbor International Airport up to 10 minutes for passengers risking a missed connection, without gate agent sign-off. The artificial intelligence system for flight delay decisions weighs schedule and transfer data, replacing case-by-case staff judgement.
Ramps Logistics — Three minutes flat: the time it now takes to quote, reserve and follow a cargo shipment moving from China, India, the US or Brazil, thanks to Ramps Logistics' new AI freight platform, CUBBY Cargo. Engineers built the system in Trinidad and Tobago, wiring it straight into WhatsApp so users chat their way through bookings instead of opening a separate website or app. That makes it, by the firm's account, the region's first AI-enabled cargo booking tool of its kind in the Caribbean.
Gazpromtrans — A 20% jump in arrival-prediction accuracy on certain routes is the headline result of a new AI in railway logistics model, built by Gazpromtrans with SigmaDevs to forecast railcar availability at stations. Station readiness for suitable rolling stock is now calculated automatically instead of by hand, freeing operators to schedule activity more precisely and pick cheaper railcar deployment options. The goal behind this AI-based rail freight forecasting: keeping production running without gaps.
Land Transport Authority — Singapore's Land Transport Authority is piloting AI bus lane enforcement, mounting onboard bus cameras to perform AI-powered bus lane detection whenever vehicles drift into reserved lanes during service hours. The scheme grew from a 2024 proposal call and sits alongside other electric-bus safety additions, among them 360-degree collision warning and anti-fatigue monitoring.
Plub — Plub, a Paraguayan grocery platform, runs AI-powered logistics for grocery delivery via Shipsy's transportation system, hitting 30-minute drop-offs. An order-clubbing engine adds AI-driven delivery routing, lifting rider output and cutting fuel use. Warehouse software targets 100% inventory accuracy and less waste, company-stated goals, not independently verified.
ČD Cargo — First nationally: ČD Cargo adopted SAP Joule for AI software development, shortening Fiori app timelines via a built-in AI coding assistant within its SAP setup. Test automation, synthetic data generation and fewer manual coding tasks followed, with the carrier crediting a 50% productivity boost, unverified independently.
SynCab (Chung Shing Taxi Ltd.) — Hong Kong's first AI-powered taxi dispatch system is now running, a joint venture between Chung Shing Taxi Ltd's SynCab fleet and the University of Hong Kong. Branded the Smart Dispatch Decision System, it pairs riders with cars by crunching live congestion levels, weather patterns and EV charge status. Drivers logged into this AI cab dispatch system cannot decline whatever booking gets pushed to them.
MTR Corporation — MTR Corporation runs two AI-powered systems handling train deployment and AI crowd control at major Kai Tak Sports Park events, and they are already dispersing passengers. A ridership prediction model built with Hong Kong University of Science and Technology on billions of MTR and government data sets joins an intelligent crowd diversion system.
Expreso Brasilia — Fatigue, distraction and speeding trigger instant alerts across more than 500 buses in Expreso Brasilia's commercial and special-service fleet, fitted with ADAS and Driver Status Monitoring hardware under a new AI-based driver monitoring in transportation programme. Satellite tracking spans the same fleet at this major Colombian inter-departmental carrier, while a ChatBot and VoiceBot now field enquiries through its contact centre and terminal kiosks. These moves anchor a broader digital transformation drive built partly on driver fatigue detection systems.
Air New Zealand — Companion AI, Air New Zealand's enterprise build of ChatGPT delivered through direct work with OpenAI, now reaches every one of its 3,500 corporate staff for daily tasks. That milestone in AI in corporate airline operations follows a year in which employees assembled over 1,500 Custom GPTs for in-house needs, with executives weighing generative AI for airline workflows across self-service for customers, upkeep scheduling and broader operational calls.
Saudi Data and Artificial Intelligence Authority — Saudi Arabia's Data and AI Authority is running a self-driving taxi pilot across seven Riyadh sites, among them King Khalid International Airport and Princess Nourah University, for 12 months. Built with the Ministry of Transport and private partners, the robotaxi scheme uses AI navigation and traffic sensors, plus an onboard safety officer.
East Japan Railway Company (JR East) — East Japan Railway Company (JR East) unveiled Japan's first AI voice assistant for train information: a spoken rail-status hotline named Dokotore Dial. Naming a station or route by phone lets generative AI retrieve live delay, departure-time and position data from JR East's train information system, aiding smartphone-averse riders anywhere.
Air France-KLM — Air France-KLM's Gen AI Factory with Accenture on Google Cloud, an instance of generative AI in aviation, cut AI build time over 35%, the partners' own figure. RAG-based tools spot damage and find spare parts, plus a staff assistant. Agent building, machine learning and model trials complete this airline AI deployment.
Nowaste Logistics — Cognibotics has installed an industrial robot at Swedish firm Nowaste Logistics' Helsingborg warehouse, an AI robotics deployment in warehouse picking built for high-volume handling. Running on Sics AI's Robot Brain One system, it reaches 2,000 picks per hour without losing accuracy and recognises unfamiliar products instantly, needing no manual reprogramming and enabling automated order fulfilment.
HHLA TK Estonia / FERNRIDE — FERNRIDE holds the inaugural EU Machinery Directive clearance covering driverless port operations, following inspection by TÜV SÜD and sign-off from the Estonian Transport Administration. Three unmanned tractors work today at HHLA TK Estonia's Muuga site near Tallinn, reflecting a wider industry move toward robotic cargo-handling machinery.
DHL Supply Chain — A £550m spend funnels 1,000-plus robots into DHL Supply Chain's UK warehouse robotics deployment, pushing its machine count from 2,000 past 3,000 beside staff. Locus Robotics and 6 River Systems units span 18 sites, and a Boston Dynamics Stretch robot debuts for container unloading, clearing up to 700 boxes hourly. Logistics automation grows nationwide.
SBB (Swiss Federal Railways) — SBB Chat, Swiss Federal Railways' AI chatbot for customer service, resolves close to 60 per cent of website queries unaided. Running on OpenAI's ChatGPT, it covers tickets, timetables and other services, passing tougher matters to 450 staff. This AI-powered rail support tool supplements rather than substitutes for that team.
Neste — Finland's Neste now runs AI-assisted shipping optimisation across Europe, built on Seaber's tool. Cargo from Finland has used the AI shipping scheduling system since 2021, cutting bunker and port costs up to 4% and emissions by a similar share via smarter fleet use.
Tashkent city government / Road Patrol Service — Uzbekistan's road patrol service is running AI traffic management from Tashkent's Situational Monitoring Centre, expecting congestion to fall by 50 per cent. The intelligent transport platform reads live vehicle flow, anticipates bottlenecks and pulls in street-camera feeds to catch crashes and violations without human input, quickening emergency response too — a benefit officials themselves project rather than an audited result.
SNCF — 100,000 of SNCF's 290,000 employees can now tap SNCF GPT, an AI chatbot for employees built on Mistral AI, and 15,000 log in daily. It draws on 70,000 legal, HR and technical records. 600 developers also use Mistral's coding software, a generative AI assistant for staff.
FlySafair — South African budget carrier FlySafair now offers 24-hour booking through Lindi, a no-cost AI travel agent on WhatsApp. The assistant handles single-passenger flight bookings, seat and name changes, plus travel queries anytime. FlySafair calls it the country's first such AI-driven chatbot.
CTP Iași (Compania de Transport Public Iași) — Romania's Iași transit fleet now runs AI anticollision system public transport safety tech across 140 buses, installed by CTP Iași. Cameras spot nearby walkers, riders and cars sitting outside driver sightlines, triggering instant alerts by sound and screen. A linked driver monitoring technology buses module also checks for phone handling, drowsiness and belt use. Xceza SRL built the setup for 6.15 million lei (€1.24m), paid for via the National Recovery and Resilience Plan.
Terminal 6 — Deep learning software from Spanish startup AllRead now reads railcar codes automatically at Terminal 6 (T6), a grain terminal in Santa Fe. Built with support from Argentine consultancy Generatica, the AI railcar identification tool scans wagons as they arrive, cutting out manual registration once needed for eight to ten trains daily, each hauling roughly 50 cars, at a site moving more than 80,000 tonnes of raw materials every day. Real-time, code-scanning technology like this had removed a longstanding bottleneck in port logistics.
Suministros & Alimentos — Suministros & Alimentos now relies on ToolsGroup's platform for AI demand forecasting in food logistics, doing live shipment tracking and stock tuning across Guatemala, El Salvador, Honduras and Nicaragua. Franchise clients such as McDonald's benefit from the setup, unveiled at Gartner's Supply Chain Symposium as a case study.
Taoyuan City Government (桃園市政府) — Taoyuan City Government has installed AI traffic control at 91 junctions near the Zhongli, Neili and Dazhu interchange areas. Sensors, edge processors and machine-learning models read congestion live, cutting driver delays by 19% and journey times over 5%. The system feeds Taiwan's wider intelligent transportation strategy.
Lufthansa Group — Munich has become the second site for Lufthansa Group's AI Tray Tracker, following a Frankfurt pilot, as part of its AI food waste tracking push. A handheld scanner reads returned trays at the dishwashing point, and AI sorts meals into untouched, partly eaten or fully finished categories. Caterers feed this AI tray scanning for meal waste data back into portion sizing and menu planning, aiming to shrink both wasted food and CO2 emissions.
Ministry of Transport and Logistics — Ethiopia's Ministry of Transport and Logistics and the Ethiopian Artificial Intelligence Institute built the Integrated Freight Transport Management System (IFTMS) for AI freight transport management nationwide. It gives officials one view of cargo, cutting guesswork. Backers expect less waste, with AI-driven cargo oversight replacing manual tracking.
AŽD Praha — EDITA, AŽD Praha's autonomous passenger train, began Europe's first open-track service on the 24 km Kopidlno-Dolní Bousov line in Czechia, using lidar and camera-based AI for driverless rail transport. A supervisor rides aboard pending unsupervised-operation rules. AŽD calls the domestically built train proof of readiness, unverified.
Ningbo Zhoushan Port Meishan Port Area — Ningbo Zhoushan Port's Meishan gates now run AI vision container terminal checks, matching plates, container numbers and seal numbers in one camera pass. Staffing fell from four to one per gate, with trucks cleared in under a minute. Automated recognition is replacing manual inspection across the port's logistics.
Maritime and Port Authority of Singapore — A live digital replica of the Port of Singapore, built with artificial intelligence, will be tested by the Maritime and Port Authority (MPA) from mid-2025, marking a fresh push into AI in port operations. Sensors tracking ships, terminal activity and environmental conditions feed this AI-powered digital twin of a port so planners can watch harbour conditions unfold as they happen, sharpening safety, efficiency and sustainability outcomes. Beyond monitoring, MPA intends to run oil-spill and chemical-spill scenarios through the model and lift situational awareness for staff managing port operations.
Qatar Aviation Services (QAS) — Autonomous airport vehicles are running airside at Hamad International Airport under 'Eye of the Pearl', built by UISEE for Qatar Aviation Services, MATAR and Qatar Science & Technology Park. One unmanned machine hauls luggage between aircraft and cargo bay; another moves passengers via sandstorm-resistant radar sensors. Officials call it a regional first for driverless airside transport.
PKP Polskie Linie Kolejowe — PKP Polskie Linie Kolejowe (PKP PLK) is piloting AI in railway operations, adopting an OpenAI system meant to slash hours wasted on paperwork and drafting reports. Traffic controllers may later access this AI deployment in railway operations, provided trials succeed, easing document handling across Poland's network.
Renfe — Renfe fitted 415 Cercanías stations in Spain with AI video surveillance, completing phase one of the RS3 (Renfe Smart Security Station) project. Algorithms flag fires, vandalism, overcrowding, unauthorised access and falls onto tracks, pushing live alerts to a 24-hour Renfe security centre. Of a budget exceeding 32 million euros, phase one cost 25.8 million euros; another 6.5 million will extend AI camera monitoring to 182 more stations.
Pošta Slovenije — Pošta Slovenije embedded PIA (Poštna Interaktivna Asistentka), an AI voice assistant for customer service, in its contact centre. Tuned for local speech, the large language model covers 12 Slovenian dialects, fields roughly 51 percent of calls, resolves 45 percent solo, and speeds replies by about 80 percent—figures from the operator itself, not independently checked, for this AI-powered contact centre assistant.
SriLankan Airlines — Nearly 12,000 queries have hit Yaana, SriLankan Airlines' AI customer support airline chatbot — 88 percent settled without staff help. CodeGen International built it on GPT-4 and NLP, easing this generative-AI customer service load on agents.
Djibouti Ports Corridor Road SA (DPCR SA) — Cracks and potholes on Djibouti's highway corridor now get flagged automatically under Djibouti Ports Corridor Road SA's (DPCR SA) new AI road monitoring system, built on the RapidCanvas AI platform for automated road condition detection. Instead of waiting for scheduled inspections, crews receive live alerts, which speeds up repair scheduling and helps DPCR SA target maintenance spending where it matters most. The operator links the deployment to safer travel between Djibouti and Ethiopia, part of a broader push under Vision 2035 to position the country as a regional logistics hub.
Cameroon Ministry of Transport — An AI trucking safety mandate took effect across Cameroon on 1 April 2025: Ym@ne Driver, an AI-powered monitoring system, became compulsory for freight trucks and interurban buses under the Ministry of Transport. Spanning 1,500 vehicles, it spots speeding, phone handling and drowsiness as they occur, having already caught over 32,800 violations. Camtrack, partnered with MTN Cameroon, runs this AI fleet monitoring scheme linking telecoms infrastructure with road-safety technology.
Finnair — Powered by Salesforce Agentforce, Finnair's AI customer service agent Sisu resolved 80% of chats without human intervention by December 2025, crunching almost 2 million LLM calls each month. This AI-powered customer support assistant converses with flyers in Finnish and English, with 15 more tongues on the way, trimming new-hire training duration by a quarter (25%).
Rossmann — Nearly every regional warehouse Rossmann operates in Germany has adopted AI route planning logistics, built with Graphmasters on its Nunav platform. Millions of traffic data points now feed real-time predictions and AI-driven route forecasting for order volumes. Eighteen months of building preceded this rollout.
Deutsche Bahn — More than 1,000 traveller queries have already gone through Kiana, Deutsche Bahn's AI ticket assistant installed at Berlin Brandenburg Airport. Fitted onto a column on Terminal 1's mezzanine, it draws on large language models to steer riders toward the correct fare even if they know nothing about tariff rules, replying in nine languages — an early sign of voice-driven rail customer support.
IKEA (Ingka Group) — Machines dominate IKEA Soroksár's floor: automated guided vehicles and robots fill most space at the world's largest AI fulfilment centre. An algorithm picks goods placement, cutting turnaround on Budapest internet orders, part of a wider AI warehouse automation push. IKEA says people still work alongside the machines.
Shuohuang Railway (CHN Energy) — Huanghuagang station now runs China's first intelligent shunting system, an automated freight rail shunting deployment by CHN Energy subsidiary Shuohuang Railway. Trains set routes, drive themselves and take remote commands instead of outdoor crews, combining AI, 5G and Beidou positioning. Danger levels fell 90%, adding roughly 10.56 million tonnes of yearly capacity.
East Japan Railway Company (JR East) — East Japan Railway Company's "AI commander" system targets AI rail disruption recovery, aiming to cut Shinkansen and conventional-line restoration times by up to 50 per cent, a milestone for AI-driven rail service restoration. Deployed on the Tohoku, Joetsu and Hokuriku bullet-train lines and Tokyo-area conventional routes, the generative AI tool monitors dispatcher-crew radio exchanges, checks them against maintenance manuals and two decades of past incidents, then independently drafts likely causes, response plans and recovery estimates. The 50 per cent figure is the company's own projection.
Air Caraïbes — Camille, an AI virtual assistant for customer service built by FCB.ai, runs on Guadeloupe-based Air Caraïbes' website, fielding over 800 daily traveller queries. It posts a self-reported 99.7% satisfaction score after six months live, part of a wider push into automated airline support.
Grupo Senda — Grupo Senda's Senda Citi division now runs AI-driven route optimisation for personnel transport in Mexico, hitting 98% punctuality. Dynamic routing, driver virtual assistants, smart safety cameras and internal AI agents lift efficiency by 5% and sharpen route planning by 8%. This AI in transport logistics backs 7,000 daily trips across 3,000 buses in four Mexican states.
Onest SmartLogistics — Onest SmartLogistics slashed losses 30% and shrank inventory checks from 10 to 3 days by weaving drones and analytics into AI in logistics operations. Oracle databases and OpenAI power the shift, with a generative assistant, OnestIA, in the works for staff on ChatGPT. The move fits broader supply chain automation goals.
Tashkent city government / Ministry of Internal Affairs — Tashkent's Road Safety Service rolled out AI traffic cameras at 10 pedestrian crossings to catch drivers who fail to yield on unsignalled crossings, issuing 1,700 fines within the first three days. The rollout will extend to 30 additional intersections, according to the agency's press secretary, as part of a wider push for automated pedestrian-safety enforcement in Uzbekistan. The detection system targets crossings without traffic signals, relying on computer vision to flag violations in real time.
All Nippon Airways (ANA) — Japan's ANA cut manual and regulation searching by approximately 90% with neoAI Chat, neoAI's generative AI platform, across airport, cabin, maintenance and flight operations. On ANA's secure private cloud, this AI in aviation shortens report and training material creation by 75%. A Narita Airport trial from August 2024 validated the airline AI.
c2c Railway — At Upminster, c2c Railway's Service Delivery Center now runs on OP1, an AI in railway operations platform combining live timetables, train positions and driver rosters into a single screen for controllers. CPC Project Services built the system, which spots trouble developing early enough for staff to head off major disruption—an artificial intelligence for train operations approach centralising data that was previously scattered.
Renting Colombia — Renting Colombia cut road accidents 34% in 2026 across 13 organisations in Rutas de Vida, an AI fleet safety scheme combining artificial intelligence, telemetry and data analysis. Cameras in over 2,000 vehicles flag tiredness, microsleep, distraction, phone use, crash risk and lane drift, alerting drivers instantly; injuries fell 45.3%. The driver-monitoring results are self-reported and have not been independently evaluated.
Ministry of Transport of the Czech Republic — The Ministry of Transport accepted the provisional EU approval of Tesla's FSD Supervised, a Czechia AI regulation step admitting partly automated driving to public roads. RDW, the Dutch authority, issued that clearance on 10 April 2026 under Regulation (EU) 2018/858; recognition rested on article 39. Classed SAE Level 2, the system keeps motorists alert, poised to intervene and wholly liable under these self-driving rules.
Ministry of Transport of the Czech Republic — Czechia now permits Level 3 autonomous vehicles on public roads, the Ministry of Transport confirmed. Under SAE Level 3 conditional automation, the vehicle steers within its lane, tracks traffic and adjusts motorway speed. A person must still be ready to resume control instantly, though keeping hands on the wheel is no longer required.
Renesas Semiconductor Manufacturing Co., Ltd. — Renesas Semiconductor Manufacturing has installed AI inspection robots semiconductor plant hardware at its Naka site in Ibaraki Prefecture: three ugo mini units that roam the Sub-Fab cleanroom alone, snapping pictures at checkpoints and compiling findings without staff involvement. The robotic cleanroom patrol tackles a labour crunch gripping Japan's chip industry by taking over repetitive checks.
Tensor Energy — Kyocera TCL Solar now runs AI battery storage control at Iizuka Daiichi Mega Solar in Fukuoka, handing a 1,500kW/6,000kWh unit to Tensor Energy's Tensor Cloud software, unattended, 24 hours daily. Sunlight and price forecasts drive automatic bidding and charge-discharge scheduling without staff involvement. This renewable energy management system links output prediction to storage dispatch.
Korea East-West Power — Korea East-West Power coached disclosure staff on generative AI for disclosure checks via ALIO pre-verification on in-house model EZY. Trained on guidelines and self-check items, EZY flags errors and mismatches and answers queries one-to-one.
RT-Engineering (Rostec) — RT-Engineering, part of Rostec, shrank industrial-facility estimates from several weeks to a few days by deploying AI agents for construction estimating. Accuracy reaches up to 90%, while people retain final decisions in this AI cost control.
Unistar Inks — Unistar Inks mined twenty years of spectrometry data to build an AI quality control manufacturing tool tracing colour faults to root cause. Now running at converter customer sites too, it recommends fixes and speeds diagnosis via this AI-based quality control method.
LG Innotek — LG Innotek has carried AI parts selection into all its business divisions, with software steering engineers toward fitting components, suppliers and prices. Reading a store of roughly 2 million components, the component recommendation system draws a shortlist within two hours, cuts new-product quotation preparation more than 70% and prices to better than 96% accuracy. The same drive sped defect identification up to 90% and shrunk camera-module production settings from 72 hours to under six.
JTC Corporation — JTC Corporation and Kajima Corporation have launched Singapore's first pilot of autonomous construction equipment, with excavators and compactors due for testing at the Bulim Autonomous Yard, JTC's sandbox in the Jurong Innovation District. Staff watch remotely and step in when something goes wrong, and JTC wants staffing 50% lower by giving one driver two machines. Widening the trial to driverless earthmovers, bulldozers and dumpers, comes under study for 2027.
H5 Resources — H5 Resources runs agentic AI mining tools at Mwerasandu mine in Ntungamo District, including an edge agent handling 1,300 sensor snapshots daily since February 2026. The Ugandan tin firm built its own AI in mining for IoT monitoring, document analysis and operational intelligence, earning it OpenAI Select Partner status in September 2026.
Kansai Electric Power Co., Inc. — Kansai Electric Power Co., Inc. has selected the Cognite Industrial AI and Data Platform for AI in hydropower operations across its hydroelectric business, forming the data backbone of its Next Generation Hydro Reborn Project. The system unifies plant data to support real-time field monitoring, 3D equipment visualisation, and transfer of veteran engineers' expertise through AI, reflecting wider adoption of AI hydroelectric data platform tools as utilities digitise ageing infrastructure.
Biogas plant at General Villegas, Buenos Aires province (system by AutoLink Solutions) — A biogas plant at General Villegas in Buenos Aires province now lets DIOXI run engines, pumps, the generator and digester loading — AI process control for biogas its builders call a first for Argentina. Sebastián Cuenca and Eduardo Espíndola, of AutoLink Solutions, worked on the system over eighteen months; the switch came three weeks before publication. Output stays constant at 1200 kW hourly. No outside party has tested the AI-run control claim.
Daewoo Engineering and Construction — Daewoo Engineering and Construction has switched on AI translation on construction sites across all its South Korean locations. Spoken Korean safety orders reach each labourer in their mother tongue as uttered, as screen text and synthesised voice. Trials of real-time site interpreting caught Korean speech at nearly 95% accuracy.
LG Display — AX SQUAD, LG Display's scheme for turning ordinary staff into AI builders, has already produced roughly 450 applications tackling problems workers spot in their own jobs. Around 190 of those live inside manufacturing alone, covering defect classification plus an optical simulation tool still being built and a human resources survey system, showing how far employee-built AI applications have spread beyond IT departments. Reach will grow further from the second half of 2026, when the South Korean firm folds in procurement, supply chain, customer support, finance, management, sales and marketing, extending workplace AI automation well past its original manufacturing and development base.
Ministry of Natural Resources — The Ministry of Natural Resources released two AI mineral exploration and mapping systems, AI-OreSeeking and AI-GeoMapping, at the 28th China Mining Conference and Exhibition in Tianjin. AI-OreSeeking automates the prospecting workflow end to end and has been trialled in more than 100 projects across over ten provincial-level regions; test data from it indicate that mineral prediction and evaluation once taking half a year can now be finished in one week. AI-GeoMapping runs from preliminary research to map compilation, identifies geological bodies with accuracy above 90 per cent and lifts processing efficiency by more than 50 per cent against traditional methods, and has been tested across nearly 100 map sheets at 1:50,000 in Qinghai, Xizang, Xinjiang and Fujian among other regions; the AI geological mapping system is also being promoted and applied in countries such as Morocco, Saudi Arabia and Laos.
GC FSK — GC FSK, a Russian developer, runs 84 AI agents on Kamerton, its own platform for AI agents in real estate, across three areas including personal assistants and cost estimation modules. Property-sector agents handled 14,610 work tasks at a 97.7% success rate, half inside 90 seconds.
Botswana Minerals — Botswana Minerals has outlined nine main exploration corridors across licences in the south, after an AI mineral exploration analysis merged decades of data over about 95,000 square kilometres. Sampling begins on the AI-picked prospects of highest priority, focusing corridors towards individual drill sites. The eight prospecting licences, some 7,000 square kilometres in the country's north-west, date from January 2026.
Santee Cooper — South Carolina's state-owned power and water utility Santee Cooper is working with Google to modernise load forecasting and financial planning with AI. CFO Tami Wilson said Google's WeatherNext 2 weather model should improve load forecasts, Google's Gemini Enterprise Agent Platform is speeding financial forecasting, and several hundred Gemini Enterprise licences are going out across the organisation.
Shandong Coalfield Geological Bureau — The Third Exploration Team of the Shandong Coalfield Geologic Bureau and partners deployed edge AI safety monitoring at four remote coalbed methane drilling projects in the Qinshui Basin, in Shanxi. It analyses video and behaviour data on site instead of the cloud, with real-time voice alerts where connectivity is weak. The team cites emergency response cut from hours to minutes, 80% better hazard-inspection efficiency and about 25 drilling hours saved on one project; none is independently evaluated.
Caterpillar — Caterpillar runs AI agents across its operations to modernise legacy code, generate and test software and catch defects earlier, and is introducing the Cat AI Assistant, agents that let field technicians find repair procedures and parts by voice. At the Ai4 conference, chief technology officer Jaime Mineart said the tools draw on roughly 1.6 million connected assets and 16-plus petabytes of data; the company plans $100 million to retrain 118,000 employees over five years.
BME — BME, an Omnia Holdings company, applies AI drill-and-blast optimisation and data analytics through a mining-specific 'vertical' AI model covering survey and planning to blast design, drilling, loading and hauling. Built into Xplosmart with BlastMap and the Xplolog drilling-data system, it uses digital twins and machine vision to follow rock from blast to crusher.
NuScale Power — NuScale Power is deploying nuclear-specific AI tools from Nuclearn and NPX for engineering and knowledge management in its reactor programme. An initial proof of concept showed they could cut time spent finding relevant information by up to 80%, with every answer traced back to its original controlled source, the company says.
Sundt & Zachry — Sundt Construction and Zachry Construction have Bedrock Robotics' Operator retrofit kit, which turns ordinary excavators fully autonomous, working on active US infrastructure sites, including a Nevada water-treatment plant and a 1.2-million-cubic-yard civil earthworks project. After about a year of supervised testing, the machines now run with no operator on site. Bedrock's chief executive, Boris Sofman, says the system sees 360 degrees and can compress schedules.
CHN Energy — CHN Energy runs 56 autonomous mining trucks among about 500 manned vehicles in shared traffic at the company's open-pit coal mine in China's north-west, coordinated by CiDi's MetaMine system. CiDi says its self-driving haulage system MetaMine had reached nearly 40 mines worldwide by mid-2026.
ExxonMobil — ExxonMobil has found four new exploration opportunities in Guyana's Stabroek Block through AI oil exploration and advanced subsurface methods. Models trained on its past discoveries, drilling outcomes and subsurface data were turned on hard-to-evaluate acreage, blending advanced seismic imaging, high-performance computing, machine learning and deep learning. Exploration vice-president John Ardill says the company is widening their use for seismic analysis and prospect generation.
Suffolk, Hensel Phelps, Burns & McDonnell — Suffolk, Hensel Phelps and Burns & McDonnell are putting jobsite AI to work. Hensel Phelps reported saving $342,000 using reality-capture tool Track3D for progress tracking on San Francisco International Airport's Courtyard 3 Connector; Burns & McDonnell vets preconstruction drawings with AI, and Suffolk's AI systems capture consistent jobsite data.
Vale — Vale and ABB have widened their alliance to take automation and AI in mining across Vale's Brazilian iron ore processing plants, as already done at Conceição II in Itabira. That 11.2Mt-a-year complex has over 7,000 automated instruments, 100-plus monitoring cameras and data intelligence tracking 400-plus variables. Operations vice-president Carlos Medeiros reports 25% higher productivity, 40% more premium direct-reduction ore and 26% lower iron losses to tailings there.
Ministry of Water and Irrigation — Jordan's Smart Transformation Project for Water and Wastewater Services Management applies AI water leak detection to its networks, ministers Raed Abu Saud (Water and Irrigation) and Sami Smirat (Digital Economy) were told at the Dabouq Operations Centre. Abu Saud called it the Kingdom's first such project, with pilot areas showing lower pressure, smaller losses and better distribution. AI analysis of consumption and demand lets the Water Authority of Jordan and water companies detect supply imbalances sooner and move affected areas onto alternative sources.
Empresa Electrica de La Habana — Empresa Electrica de La Habana will hand dispatchers AI blackout scheduling tools built with mathematicians at the Facultad de Matematicas of the Universidad de La Habana. The utility says the decision-support software, shown to President Miguel Diaz-Canel, should even out how circuits are rotated and push notices to households over Telegram. Nothing is switched on yet: the company describes AI-driven outage rotation as work still under way, aimed at transparency and fairer sharing of cuts.
Novo Construction — Novo Construction is using Diffs, BuildCheck's AI tool, on two California projects to flag design changes between successive drawing packages pre-construction, replacing manual review of hundreds of pages. CIO Colin Stoner says it helps teams price early and track evolving drawings, while urging 'trust but verify' on AI output.
ADNOC — ADNOC and SLB run a DrillOps-powered Real-Time Operations Centre across 120-plus onshore and offshore rigs, feeding live data to AI drilling analytics that flag problems early. ADNOC credits the AI-led monitoring with 30 to 40 per cent less engineering effort, two to three times more rigs per engineer and up to two days' downtime avoided; no independent evaluation is published.
Eskom — Eskom runs 20 AI initiatives across eight business areas — customer engagement, predictive maintenance, cybersecurity, energy trading analytics, workforce optimisation, AI infrastructure — plus 220-plus pilots, part of its AI in energy utilities push. Drones survey its 33,369 km grid, aiding repair crews: AI-powered grid maintenance underway.
İSKİ (Istanbul Water and Sewerage Administration) — Istanbul's water and sewerage administration, İSKİ, launched the in-house İSKİ AI platform on 20 July for AI consumption forecasting, fault prediction and handling citizen requests. Deputy mayor Nuri Aslan says the Alo 185 smart assistant will cut call-centre workload by about 40% and forecasts will reach 97% accuracy.
Tokuyama Corporation — Slashing yearly travel costs by 5% and cutting scheduling time to a quarter, Grid Inc.'s ReNom VESSEL delivers AI ship scheduling to 40-plus Tokuyama Corporation cement sites nationwide. Forecast weather drives route calculations, a first among Japan's coastal carriers, via this AI-driven vessel route optimisation.
Niger Delta Power Holding Company — Nigeria's Niger Delta Power Holding Company runs predictive maintenance AI in power generation, tracking gas turbine vibration, heat and part wear to flag deterioration before an unplanned shutdown. This AI-driven turbine failure prediction replaces fixed-schedule servicing with condition-based alerts.
Ethiopian Electric Utility — Preventing blackouts nationwide now hinges on an AI power grid rollout by Ethiopian Electric Utility (EEU). Vulnerable sections are flagged automatically, and emerging faults get diagnosed on the spot, swapping reactive fixes for predictive fault detection. Getu Geremew tied the shift to fresh handheld kit for field crews plus a Huawei tie-up modernising backbone communications.
Daihatsu Motor Co., Ltd. — Daihatsu Motor's Shiga (Ryuo) plant now catches hairline scratches inside machined transmission holes via an AI quality inspection system built with VRAIN Solution. Cameras and pattern recognition replace eyes trained by years of shop-floor intuition — machine-vision-based defect detection taking that call from people.
AS Augstsprieguma tīkls (AST) — Latvia's power grid operator AS Augstsprieguma tīkls (AST) is trialling AI infrastructure inspection to shrink manual review time by 30 to 50 per cent. Partnering with SIA Datacom under EU funding, AST trains a prototype to spot damage across thousands of drone-shot images of transmission infrastructure. This aerial grid-defect detection tool remains at pilot stage.
Toyota Motor Corporation — 436 Geekplus AMRs move parts unattended across Toyota Motor Corporation sites in Japan, roughly 200 per installation, marking this autonomous mobile robots Japan plants rollout. Coverage spans receiving to picking and processing, a robotic material handling swap for hand-pushed carts amid staffing gaps.
Aramco — Aramco has plugged Emerson's AI-driven Aspen Hybrid Models into its refinery planning, which Emerson calls among the largest multi-site, multi-period optimisation models anywhere. In Continuous Catalyst Regeneration and Platformer units, this AI refinery planning supports feedstock blending and margin forecasting, with yield and quality predictions up to 98.5% accurate in key units, per Emerson. Aramco says the hybrid models cut engineers' manual adjustments, and hydrocracker units come next.
Marunouchi Heat Supply Co., Ltd. — Live-equipment trials of AI plant automation are underway at Marunouchi Heat Supply's Otemachi sub-plant in Tokyo, running Preferred Networks' PlantPilot software. Trained on past operating records and operator judgement, it forecasts conditions and proposes control responses. Marunouchi's push toward AI-driven district heating targets full automation by 2027.
KoBold Metals — KoBold Metals, bankrolled by US billionaires Jeff Bezos and Bill Gates, has begun what the firm bills as history's biggest AI mineral exploration drive: 13 permits spanning 3,000 sq km of the Democratic Republic of Congo's Manono belt, set to widen to 5,000 sq km by December. Over $50 million is committed through early 2027, funding aerial coverage of 30,000 sq km and beyond 30,000 soil assays. Machine-guided prospecting redirects crews weekly towards the strongest lithium ground, and the MetaLab container laboratory pulverises, dries and tests samples on site within days. Chief executive Kurt House says a year back the outfit owned no acreage and employed nobody there; it now ranks as America's biggest investor in the country.
Elsewedy Electric — IBM is powering an agentic AI enterprise deployment at Elsewedy Electric, an Egyptian energy solutions provider, via watsonx.ai and watsonx Orchestrate. Over 30 use cases span finance, legal, human resources and market intelligence; more than 10 already run at advanced stages, yielding measurable productivity gains, the company says. This enterprise-grade AI automation replaces scattered trials with governed, action-oriented workflows.
ČEZ — ČEZ has switched on AI predictive maintenance for the turbine at its Štěchovice II pumped-storage station, calling the installation among the newest instances of such technology entering power generation. Microphones and vibration and pressure probes sampling at tens of kHz sit in tightly specified spots on the machine, feeding roughly 2,000 GB of condition monitoring data every month; the Faculty of Mechanical Engineering at Brno University of Technology co-developed it. Róbert Heczko, in charge of the group's hydro fleet, said findings to date show some impending faults could be caught far ahead, letting outage planning improve — a reading the company has not had independently evaluated.
KoBold Metals — KoBold Metals has started AI mineral exploration over Burundi's geoscientific archives, a three-year job commissioned on 12 March 2026 and paid for by the firm. Holdings from national services and the Royal Museum for Central Africa are going digital, targeting scandium, nickel, copper, cobalt and platinum group metals. The government keeps ownership, local technicians are in training, and some material was due on the web by July 2026, with machine-learning prospecting sifting the files.
Bellegprom — Belarus's Bellegprom has AI clothing design running at member factories, with the output already on shop shelves. Svitanka knitwear carries appliques and prints drawn that way; Belvest applies the same approach to footwear. Olga Grabenya, the concern's trade and marketing head, says such stock moves through Bellegprom's own branded network. Designers and technologists stay on, because AI-generated prints still need someone judging what shoppers want.
Toyota Motor Manufacturing Canada — Toyota Motor Manufacturing Canada has signed a commercial deal with Agility Robotics to run Digit humanoid robots in its Ontario vehicle plants, after a year-long pilot with three robots through development, proof-of-technology and onsite phases. Seven more robots will move totes on and off an automated tugger. Toyota's plant president cited team member experience and operational efficiency.
Minera Chinalco Perú — Minera Chinalco Perú has finished rolling out its Integrated Operations Management Centre (GIO), an AI mine operations centre heavy on analytics that now runs the entire Toromocho copper mine production chain under a hybrid organisational model. Staff are being trained in analysis, automation and remote operation.
Saudi Water Authority — The Saudi Water Authority and IBM have launched the H2O Platform, generative AI and hybrid cloud for AI water management across over 40 use cases in the Kingdom's water value chain. The water-resource AI offers live monitoring, trend spotting and predictive insight, per IBM, and a Consumer Assistant gives the public water services, consumption tracking and conservation tips.
SOCAR Carbamide — SOCAR Carbamide at Sumqayit, Azerbaijan, has joined the World Economic Forum's Global Lighthouse Network among 22 freshly named factories, recognised for industrial AI in fertiliser production. Squeezed from 2022 by broken regional food routes, shaky natural gas feedstock and already flat out, the works layered 42 digitally enabled use cases over an in-house AI-machine learning engine, adding closed-loop autonomous process control, generative AI and shopfloor robots. Throughput climbed 21% and gas efficiency 24%.
Cementos Argos — Cementos Argos has swapped manual grinding and inventory planning at its Cartagena plant for machine learning in cement production, an optimisation system that weighs hundreds of scenarios simultaneously and proposes actions. Built by in-house planning, production, supply-chain and analytics teams, it will spread to other Argos plants.
Sirte Oil Company (Libya) — Sirte Oil Company doubled its rate of penetration over earlier Al-Khair field rates with Libya's first AI directional drilling, working with SLB, the National Oil Corporation (NOC) says. An AI reservoir advisor read petrophysical logs live; paired with a drilling advisor, it drove automated drilling and geosteering, keeping the well within the target layer.
Turkmenbashi Complex of Oil Refineries (TCOR) — Turkmenbashi Complex of Oil Refineries has finished modernising, adding automated production management, live SCADA monitoring, ERP platforms and AI in oilfield operations through Smart Fields and Smart Wells, which interpret geological data to lift well performance. Seydi Oil Refinery has added digital monitoring for predictive maintenance. Turkmengeologia State Corporation reports both operational, with AI-driven well optimisation cutting costs and improving resource management.
Strix Plc — Isle of Man-based kettle controls maker Strix Plc expects AI-Barbara, a machine learning tool at its Ramsey factory, to add 4.3 million components and save approximately £50,000 every year. Bringing AI in manufacturing to tooling, the system reads real-time production data to forecast the best settings, turning set-up times of up to four hours into right-first-time runs. Local consultancy Insight Innovation wrote the TensorFlow neural network, a manufacturing AI funded through the government's Activate AI programme.
Government of Gabon — Gabon added Forest Foresight, co-developed with WWF, to its national forest surveillance system, using AI deforestation prediction to flag at-risk areas up to six months ahead with 80% accuracy. This satellite-based deforestation forecasting has already assisted in stopping an illegal gold mine threatening 30 hectares of primary forest.
Elektroprivreda Srbije (EPS) — Elektroprivreda Srbije (EPS), Serbia's state utility, runs an AI inflow prediction hydropower system for the Drina-Lim basin, roughly 10% of national electricity. Built with the Institute for Artificial Intelligence and Mihajlo Pupin Institute, it uses weather and river data for machine learning based inflow forecasting, guiding maintenance and generation planning.
SafeSip — Rural Tanzania now gets round-the-clock AI water purification via SafeSip's solar-powered "water banks" — AI-enabled water filtration units that run unstaffed. Built-in algorithms flag clogging filters, warning technicians before failure. Each 20-litre container costs 250 Tanzanian shillings.
YPF — YPF opened its Sala Integrada control room at Plaza Huincul carrying five AI agents in oil and gas operations, built to answer real-time technical questions from staff running the complex. The room joins the Real Time Intelligence Center with the plant's existing control room, marking Argentina's first such merger, so operators can optimise output using an AI-driven control room for industrial production. YPF's own projection puts the added margin above US$4 million within year one, a company estimate not yet independently audited.
EDF Martinique — EDF Martinique applies AI for power grid diagnostics along a 40-year-old high-tension line running between Le Marigot and Le Lorrain, pinpointing its most vulnerable stretches for priority repair. Drones supplement this scan, catching faulty cables under the utility's PDV-HTA scheme, an AI-based power line maintenance approach built to lengthen the ageing grid's service life instead of a full rebuild.
First Quantum Minerals — First Quantum Minerals has boosted output by over a third at Sentinel, its copper mine in Kalumbila, Zambia, after winning regulatory sign-off for automated mining drilling built with supplier FLANDERS. Three robotic rigs with laser sensors and GPS improve hole precision and blasting results while lowering hazards. Staff now run the equipment from remote control rooms.
HL Mando — A South Korean-built humanoid, A-Robot's Alice M1, now works HL Mando's Wonju plant floor, a humanoid robot manufacturing test backed by the M.AX programme. It handles dull jobs staff dodge. First homegrown humanoid at a major domestic factory, a milestone for robotics for industrial settings.
Kenya Power (KPLC) — All eight Kenya Power regions now use OCR meter reading, following a winning six-month Nairobi pilot. Crews scan the dial and the reading uploads automatically, no keypad needed. Officials plan a wider rollout of this Optical Character Recognition system to 1.8 million postpaid meters still visited by hand.
Scania — Scania runs ChatGPT Enterprise worldwide for AI in engineering, with OpenAI. Grassroots onboarding beat adoption forecasts, embedding gains into continuous improvement via bottom-up experimentation — a model for AI for engineering teams elsewhere.
Ennatuurlijk — Ennatuurlijk has widened its Gradyent partnership to bring a digital twin district heating deployment onto Midden West Brabant, its biggest network, serving Breda and Tilburg with about 1 TWh of heat each year. Software from the Gradyent Digital Twin will handle output levels and temperature settings as the site moves off biomass towards greener sources by 2027. Elsewhere in the Netherlands, this Dutch provider keeps 62 separate networks running for 90,000 households and 1,200 firms, a heat network optimisation effort now spreading beyond one site alone.
Tapestry consortium (Northpower, Orion Group, Unison Networks, WEL Networks) — Northpower, Orion Group, Unison Networks and WEL Networks now run GridAware, Tapestry's AI grid inspection platform, spanning New Zealand networks with over 510,000 customers — the nation's first such grouping. Drones and computer vision replace manual surveys, spotting faulty or ageing gear. Within two years the four aim to supply 10,000-plus labelled images, powering automated infrastructure monitoring.
Taisei Corporation — OpenAI's first Japanese general-contractor partner, Taisei Corporation, is running firm-wide ChatGPT Enterprise adoption to retrain staff and reshape operations, a leading example of AI in construction industry deployment. From 250 users in April 2025 to 1,000 by August, results showed 5.48 weekly hours saved per person, 6.6万 hours saved yearly overall, 4.20 satisfaction, and 4.51 willingness-to-use. Generative AI in construction now heads for company-wide expansion.
Guam Power Authority — Guam Power Authority will deploy AI weather forecasting for its electric grid through the ARMOR project, backed by $1.62 million in US federal grants. A Total Sky Imager uses machine learning cloud cover predictions to help operators adjust generation in real time and respond to storm-related outages.
Hoff SA — Digel's generative AI potato processing tool is being trialled at Norwegian cooperative Hoff SA's Gjøvik site. Staff query the line in plain language, pulling live data and manuals for fixes — among the sector's first such factory deployments.
Entek Elektrik Üretimi and KoçDigital — Entek Elektrik Üretimi and KoçDigital hit 90% water flow forecast accuracy with a fresh hydroelectric plant optimisation tool, boosting energy production efficiency up to 2.5%. Crunching 56 parameters spanning 25 years of past records, it sharpens dam output scheduling nationwide. AI-based hydropower forecasting now replaces manual guesswork.
YPF — YPF teamed with Globant to roll out Digital Suppl.AI, a supply chain AI platform powering 46 agents across eight agentic solutions within the Argentine state energy firm. Conversational agent handling now covers purchasing and stock oversight at the company, retiring manual routines in favour of live strategic guidance. Wider deployment follows this opening stage, which centres on AI in procurement and inventory management.
Aramco — Autonomous AI agents gas plant control now runs at Aramco's Fadhili Gas Plant, jointly engineered with Yokogawa Electric via a reinforcement learning algorithm overseeing acid gas removal. Amine and steam consumption dropped 10 to 15 percent, power draw shrank 5 percent, and staff seldom step in by hand anymore. Saudi Arabia's facility ranks among the earliest large-scale adopters of self-governing process control.
Emirates Nuclear Energy Company — Emirates Nuclear Energy Company (ENEC) now runs ENECIQ, an agentic AI platform built with Presight, across corporate systems in the United Arab Emirates, drawing on ENEC-specific knowledge bases for custom AI insight generation, condensed technical reports and automated routine work. It stays inside ENEC's secure cloud, apart from nuclear reactor operations.
Erdenet Mining Corporation — Erdenet Mining Corporation runs 100% automated bid scoring, no human review needed, via AI tender evaluation software TenderEval, made by Chimege Systems. Mongolia's biggest state copper firm fields nearly 1,500 tenders yearly; the tool spares reviewers from combing hundreds of pages by hand while widening openness in buying, an automated procurement scoring push dated 2025.
Pertamina Hulu Rokan — 12,600 wells across 6,400 km2 in Indonesia's Rokan Working Area fall under AI-based well management from Pertamina Hulu Rokan (PHR), a Pertamina unit. Its Digital & Innovation Center (DICE) pushes live data onto 66 screens spanning drilling, output, upkeep and shipping, flagging quick suggestions. This AI oil well monitoring sits inside the OPTIMUS programme, chasing up to USD 46 million in savings by end of 2025.
CMPC — CMPC, the Chilean pulp and paper company, saves 13 minutes per maintenance note with voice recognition, part of a rollout of AI agents in manufacturing across mills in Chile and Brazil for industrial process automation. A linked tool turns two-hour planning reviews into ten minutes; at Santa Fe, production KPIs are watched live, with alerts sent to floor staff via WhatsApp or email.
Repsol — Repsol runs a new multi-agent AI platform, built with Accenture, spreading autonomous AI agents — more than 20 of them — across three teams in its Digitalisation and Services division of roughly 700 staff. Over 50 employees supervise these self-organising bots as they coordinate and finish jobs, drawing on knowledge bases made solely for the Spanish energy firm. A custom interface fits large language models to this workplace after a four-month trial.
RAK Ceramics — RAK Ceramics runs self-driving electric trucks at Al Hamra Industrial Zone, Ras Al Khaimah, the UAE's first such commercial fleet, with Evocargo. Fifth-generation autopilot and an AI multi-sensor system guide the Evocargo N1 vehicles between plant sites, up to 200 km per charge, pausing only to load, unload or recharge — an early case of autonomous trucking in the region.
Kenya Power — Kenya Power built Nuru, an AI chatbot Kenya Power says will serve 10 million-plus customers via its website and Facebook page. Nuru fields outage reports and routes harder questions to staff, arriving with a refreshed MyPower App and Kiswahili option on USSD *977#, extending AI customer service in Kenya's power sector.
Cargill — Cargill's Amsterdam site, a multi-seed crushing and refinery facility, is trialling Spot, Boston Dynamics' four-legged robot, for around 10,000 weekly AI safety inspections in agribusiness. Thermal, acoustic and visual readings flag overheating machinery, product leaks. Cargill calls the AI-driven inspection pilot a step towards predictive, digital-first plant management.
Xingfa Group — A single control unit now runs autonomous chemical production at Xingfa Group's chlor-alkali site in Hubei, built by Supcon under the label industrial embodied AI. Rather than relying on separate control cabinets, the plant assigns everything to what it calls its 81st digital employee — software combining Supcon's TPT time-series large model with the firm's UCS platform to monitor over 15,000 data points and act on them without human input. Results so far include caustic soda held steady between 32 and 32.1% concentration, plus valve-fault diagnosis accuracy topping 95%, an early sign of embodied AI reaching chemical manufacturing.
Midea — World Economic Forum Lighthouse Factory status now belongs to Midea's Chonburi plant, the first such honour for an appliance maker outside China. The 208,000 sqm Thailand site runs Southeast Asia's first fully 5G-connected line, joining 5G to AI in manufacturing so cameras catch faults as goods pass through, an AI-driven manufacturing intelligence approach. Power use sank 40.2%, water 23.5%.
GlobalFoundries — GlobalFoundries' 300mm Singapore fab won Global Lighthouse status from the WEF (World Economic Forum) after rolling out over 60 AI in semiconductor manufacturing upgrades since 2020, spanning ML, IoT and analytics tools. Labour productivity rose 40 per cent and prototyping time for new products fell 30 per cent, as machine learning took over upkeep forecasts, defect screening and scheduling tasks.
Kuwait Oil Company — Agentic AI drilling now runs at Kuwait Oil Company's rigs, the debut win for its Artificial Intelligence Innovation Center. Built on Ghaia.ai's G Agent platform with Microsoft and Halliburton, it fields agents Ahmad Al-Eidan says think, act, and evolve, aiding planning and real-time data analysis. This is Kuwait's first large-scale use of autonomous AI agents.
MVM Group — Tens of thousands of customers relied on Rezsiőr, MVM Group's freshly launched chatbot, during its debut month to sort out queries on subsidised tariffs, gas fees and power charges. Crafted by Hungarian firm YoloAI and fitted to MVM's billing infrastructure, the tool illustrates AI customer support for energy bills operating at national reach. MVM, Hungary's state-owned utility, positions the AI-powered utility billing assistant as backup for staff, not a replacement.
Tokuyama Corporation — Tokuyama Corporation extends Cognite Data Fusion and Cognite Atlas AI to nine Tokuyama Factory departments by March 2026, merging sensor data, equipment diagrams, maintenance systems and inspection records. Japan's chemical and materials manufacturer uses AI in chemical manufacturing to automate operations, improve anomaly detection and speed inspection tasks: digital transformation via chemical plant AI.
K-Print — K-Print, an Estonian printer, automates a third of orders via an AI in print manufacturing system built for a large export deal. This automated order-processing shift expanded capacity without matching staff growth.
GL Group — Onshore fields in Salyan now run AI seismic data processing under a Waverity partnership with GL Group, marking a regional first for post-stack interpretation of subsurface data. Kursangi and Garabaghli saw sharper imaging and newly identified zones missed by earlier methods. GL Group, based in Azerbaijan, showcased results from this machine-learning approach to seismic mapping at Houston's IMAGE 2025 conference.
National Grid — National Grid now runs AI-powered drone inspection across pylons and cables in England and Wales, billed as the first centralised system worldwide. Pilots in one control room fly sees.ai aircraft beyond direct sight, freeing engineers from routine checks. Manual line-walking gives way to automated grid surveillance.
Midea Group — China's Midea Group earned WRCA recognition for its Jingzhou washing machine plant, the first multi-scenario AI agent factory manufacturing site anywhere, a benchmark for AI in industrial production. A unifying brain pairs large models with embodied robotics, directing 14 agents across 38 core production scenarios there. Hours of hands-on labour now finish in seconds, lifting average efficiency beyond 80% and scheduling responsiveness by 90%.
PLN Nusantara Power — PLN Nusantara Power, subholding of Indonesia's state-run PLN, deployed Hydrobot, operator assistant applying AI in hydropower. Using natural language processing, live weather, telemetry and predictive models, the hydroelectric AI tool forecasts inflow discharge and dam surface elevation, optimising reservoirs against overflow. Hydrobot won the Gold Stevie Award for Technology Breakthrough of the Year (energy technology).
GS Engineering & Construction — GS E&C is first among South Korean builders to adopt ChatGPT Enterprise, spreading AI in construction into contract checks, drawing analysis, cost forecasting and technical review. OpenAI staff visited Seoul to demonstrate embedding the tool into site workflows. A GPT Champion scheme from September trains staff as specialists to widen construction sites AI use cases.
Endesa — Endesa now applies AI in energy operations spanning Generation, Distribution, Commercialisation, Tax, Human Resources and Customer Service. Predictive maintenance and automatic image diagnosis back power generation; drones, sensors, fraud detection and grid monitoring cover distribution. Generative AI drafts personalised energy-saving proposals and reads customer emotions live, while virtual assistants handle tax and HR work.
Rhino Mabati Factory Limited — Kenya's Rhino Mabati Factory Limited leads the construction-materials sector in AI-powered robotics in roofing production, run on live plant lines. Machine-learning-steered robotic arms shape corrugated sheets, part of wider AI-driven manufacturing processes also used at a new gypsum-channel plant. Offcut waste has reportedly fallen up to 25 percent, an unaudited company figure.
Fonterra Co-operative Group — Every one of Fonterra's 66 million annual milk powder bags now passes under AI image recognition milk powder inspection, part of a system covering all 56 packing lines. Designed in-house, this AI visual inspection setup catches damage on the 25kg sacks, pulls faulty ones off the line instantly, and stores a timestamped photo of each for tracking back through production. Separately, New Zealand's dairy giant has been trialling the Plant Equipment Scheduling Programme (ESP), which applies AI-driven optimisation and is now spreading from Edgecumbe toward Clandeboye.
Gold Fields — Gold Fields ran a 10-month AI safety pilot for mining at its Cerro Corona site in Cajamarca, now certified by Chile's CIPTEMIN. SafetyMind's computer vision platform scans existing cameras, flagging unsafe conditions in real time. This makes Gold Fields the first Latin American miner with an independently verified AI preventive safety system, under its Seguridad 4.0 strategy.
Sucroal — Sucroal, a chemical and biotechnology firm in Colombia's Valle del Cauca, uses AI document processing to clear each export file in ten seconds, down from seven minutes, saving roughly 40 hours monthly across 350 files. Beyond paperwork automation, AI starts over 50% of new software development; plant-floor computer vision counting and machine learning for Aspergillus niger strain improvement are in testing.
Sinopec Shengli Oilfield — 40 subsidiaries of Sinopec's Shengli Oilfield now share China's first large-scale AI in oil and gas system, the Sheng Xiaoli large language model. By August 2025, 1.8万 accounts had signed up, sending 2,000-plus daily queries for document analysis, data retrieval and report drafting. A DeepSeek-R1 upgrade brought 280 assistants and 1,000-plus knowledge bases into this petroleum-sector generative AI platform.
After You — An AIoT factory energy management deployment at After You's Samut Sakhon plant, built with PTT, cut power spending 15% via hands-off cold-room thermostat regulation. Linked IoT sensors and AI algorithms watch chiller output round-the-clock, driving an APA (Automation Process Analytic) setup for AI-driven cold room control that reacts on its own overnight, unstaffed.
QazaqGaz — Manual home visits are disappearing from gas billing at QazaqGaz Aimaq, Kazakhstan's biggest gas supplier, now trialling AI meter reading via an app that reads consumption straight off a snapped photo. Kyzylorda, Kostanay and Turkestan host the pilot, aimed at quicker bills.
Sangan Steel Pelletizing — Iran's Sangan Steel Pelletizing began testing AI in steel production in 2025 via its SMARTizing scheme, mining historical output figures to anticipate porosity and cold crushing strength before they occur. Ideal settings emerge through machine learning, shrinking energy and raw material demands. This steel industry AI effort stays pilot-stage, not yet fully rolled out.
Santos — Papua New Guinea's wells, pipelines and reservoirs now draw on Xecta's Integrated Production System Model (IPSM), an AI-powered digital oilfield system Santos rolled out after success in Australia's Cooper basin. Physics-based modelling paired with AI-driven production optimisation converts billions of telemetry readings into continuous, live output guidance.
Novovoronezh Nuclear Power Plant (Rosatom) — Rosatom's Novovoronezh Nuclear Power Plant is testing an AI operator support system for nuclear power plants on unit 6, spanning 360 systems and nearly 200 procedures. Live readings, anomaly flags and 30-minute forecasts now substitute for staff scanning over 12 thousand sensor feeds — a shift toward automated reactor monitoring.
Renault Group — Renault Group's Plant Connect at the Palencia plant in Spain crunches 5 billion data points daily across 110 square metres, driving AI manufacturing decisions. Its 24 stations use 120 microphones for acoustic faults, 3D scanners for shape mapping, and vision-based defect detectors flagging flaws to 0.6mm in battery connectors. Output: 555 cars daily.
RUSAL — Aluminium melt preparation at RUSAL's Taishet Aluminium Plant now runs under a neural network built to spot faulty batches before they form. The AI machine vision aluminium production system, developed jointly by RUSAL's Engineering and Technology Centre and the plant's foundry team, watches up to 20 steps in the process, confirms the right sequence is followed, and flags instructions to staff on-screen. RUSAL intends to extend this AI-based alloy quality control approach to every smelter it operates, cutting the odds of substandard output.
Ferrostaal Chile — Ferrostaal's Chile operation now uses AI recruitment for mining through Mila, an agent that screens candidates and sorts CVs for industrial vacancies. Fed on real CV data, it parses applications, spots gaps such as missing attachments and answers candidates straight away. Ferrostaal credits the deployment with a 60% drop in initial CV handling time and a jump of more than 40% in how many profiles reviewers get through each day, pointing to a broader shift toward automated candidate screening in heavy industry.
Metinvest — Metinvest Digital, the IT arm of Ukraine's Metinvest mining and metals group, now runs computer vision AI steel quality control on semi-finished slab output, catching flaws instantly and directing workers on precise cutting points. R&D centre head Maksym Balaniuk disclosed the rollout on 20 June 2025 at a gathering in Kyiv, noting the tool already operates at the firm's biggest facility. A comparable build for swifter lines moving at 5 metres per second, an AI-based defect detection setup for steel manufacturing, remains in progress.
Pluspetrol — Peru's Pluspetrol has finished testing EXR-2, an autonomous unit built for AI gas leak detection, at the Camisea consortium's Pisco fractionation site. Thermal imaging, precision sensors and smart navigation let it spot hot spots, pick up unusual noise and send live readings to a control room, plus build a 3D map for autonomous plant inspection.
Rio Tinto — Rio Tinto's ReconAI tool powers AI rail defect detection across its automated freight network, reaching 96 per cent classification accuracy in live operation on autonomous locomotives. Pulling data from multiple systems, it flags recurring equipment problems and cuts manual admin work for maintenance staff. The mining group's fault-sorting system, a form of AI-based fault classification for rail assets, lets engineers focus on fixes rather than repetitive logs.
Rosneft — Rosneft's Tyumen Oil Research Centre built LitoText, an AI assistant for AI core sample analysis that scans geological reports for 16 physical parameters, among them rock type, colour, saturation, texture and grain size, plus fracturing. Powered by large language models, the lithological description tool cuts analysis time five to seven times over while lowering human error. It now handles fresh descriptions and audits older digital records across 52 km of stored core.
Pasargad Energy Development Company — Iran's oil and gas sector has gained artificial intelligence woven into Pasargad Energy Development Company's (PEDC) Well, Reservoir and Facility Management (WRFM) platform, paired with a freshly rolled-out Upstream AI Assistant (UIA) — described as the nation's first such tool built on large language models. Staff working the Sepehr and Jofeyr fields now get live output monitoring, forecasts guiding drilling decisions and automatic warnings, whilst UIA helps specialists retrieve technical answers from global databases rather than searching manually themselves. WRFM had already gone live back in December 2024, long before these AI-based upstream capabilities for petroleum work arrived.
Compañía de Minas Buenaventura — Compañía de Minas Buenaventura is rolling its self-built AI mining software, YupAI, out to two further sites, El Brocal and Uchucchacua, having already proven the tool at San Gabriel. Its name comes from the Quechua verb for 'to count', and it applies open-source algorithms toward AI-driven resource estimation, projecting key variables with no need for new drilling or laboratory work. That earlier deployment alone modelled 390 drill holes absent chemical assays, sparing roughly six months plus the extra expense.
KazMunayGas — KazMunayGas has deployed ABAI for AI oil exploration and production across seven modules covering wells, repairs and flood control. Flooding-event readiness fell from 40 hours to two, saving KZT450 million, with KZT9.3 billion projected by 2030 — an AI-based oilfield safety and speed upgrade using Big Data (unaudited company figures).
Coca-Cola Beverages Malawi — USD14.9 million has gone into a Lilongwe line for Coca-Cola Beverages Malawi, part of Coca-Cola Beverages Africa, built around AI-driven predictive maintenance in manufacturing. AI-powered fault detection on production lines flags problems early, cutting downtime. The 19,200-bottle-per-hour line will supply Zambia exports and staff training.
Botswana Diamonds — Botswana Diamonds (BOD) uncovered seven previously unreported kimberlite anomalies and 11 polymetallic targets by running AI kimberlite exploration on old geophysical, soil and drilling records instead of months of manual review. That mineral exploration AI push now backs BOD's filing for 11 prospecting licences over 7,000+ square kilometres in Botswana.
Mercedes-Benz — Apptronik's Apollo humanoid robots now work with AI in automotive manufacturing at Mercedes-Benz's Berlin-Marienfelde Digital Factory Campus. Apollo carries parts to the assembly line; MO360 AI Factory spots quality flaws instantly via a chatbot network and data science team, weaving humanoid robots in manufacturing into plant work.
Midland Oil Company — Midland Oil Company, Iraq's state-run operator, has rolled out AI oil well monitoring at East Baghdad, a national energy-sector first. With Chinese firm EBS, staff run the DeepSeek-R1 system for round-the-clock surveillance, cutting detection delays to minutes versus manual checks. This AI-based well surveillance beams alerts to smartphones for instant visibility.
Eurasian Resources Group — Eurasian Resources Group is applying AI in mining at Kazchrome's Donskoy chrome ore plant in Khromtau, Kazakhstan, where cameras now catch belt overheating before it starts. This machine vision for predictive maintenance approach cut overheating downtime by 75 per cent and matched human operator inspections on 80 per cent of detected flaws, with ERG's own figure crediting the rollout with over US$300,000 in cost savings within three months.
Korea Western Power — Korea Western Power trained WP-GPT on 1,432 internal documents and roughly 35,000 Q&A records, becoming the first of South Korea's state-run power generators using generative AI for power plant maintenance. WP-GPT covers maintenance, procurement, construction and safety queries, running on a fully internet-isolated network to stop data leaks. By year-end, the plant maintenance chatbot is due to draft procurement specifications and equipment management reports too.
Sonatrach — Rig-mounted cameras and live telemonitoring now feed Sonatrach's new ENTP EYES / Real Time Decision Support Center at Hassi Messaoud, an AI-powered drilling monitoring hub the Algerian state oil and gas firm has opened to track performance and streamline resource use during drilling and prospecting. The facility signals a broader turn toward AI in oil and gas decision-making, giving engineers a continuous operational picture instead of periodic reports.
Cooperativa Rural de Electrificación (CRE) — CREcencio now runs non-stop WhatsApp support for Cooperativa Rural de Electrificación (CRE), an AI chatbot for customer service fielding hundreds of outage reports, QR payments, consumption checks and bills together, 24/7. Answers rely on artificial intelligence and remembered chat history, not scripts; no purchased AI virtual assistant platform was used — CRE coded it in-house.
Energinet — A new machine learning model now drives AI grid balancing for Energinet, shaving roughly DKK 1 million a week off reserve-capacity spending. Weather forecasts and past records let the system pinpoint exact backup power needs, avoiding costly over-buying. Denmark's state grid operator recovers costs via transmission tariffs, so sharper electricity demand forecasting also eases household bills.
Cementos Cienfuegos S.A. — Cementos Cienfuegos S.A., Cuba's state-owned cement plant, now runs AI process efficiency diagnosis across its industrial operations. This AI in cement manufacturing case is one of several documented across Cuban industry.
Apollo Tyres — Manufacturing Reasoner, Apollo Tyres' in-house agentic AI platform, reads IoT data from 250-plus curing presses, cutting fault diagnosis from two hours to under 10 minutes, manual effort by 88 per cent, and costs by roughly ₹1.5 crore yearly — agentic AI in manufacturing at work. A linked GenAI tool for production issue diagnosis fields 14,000 monthly commercial calls.
CEAT — Twenty per cent higher mixer output at CEAT followed the rollout of an AI-governed closed-loop control system, one strand of a wider AI in tyre manufacturing push across the Indian firm's plants, alongside an 18 per cent drop in cycle time variation. Cameras and computer vision now flag inspection faults automatically, machine learning schedules upkeep before breakdowns occur, and a smart export container optimisation tool has slashed order-to-dispatch turnaround by 54 per cent. Support queries flow through chatbots, while designers trial tread patterns shaped by an AI-driven manufacturing supply chain approach.
Ashok Leyland — Hubble.ai now ties together 28 assembly lines and 2,000 machines inside Ashok Leyland's plants, forming one AI manufacturing platform built entirely in-house rather than bought off the shelf. Cameras under the Manufacturing Process Assurance System (M-PAS) catch defects as they happen, while predictive maintenance and robotic process automation (RPA) keep equipment running; a bot named ALPHA clears more than 10,000 supplier bills daily on its own. Ashok Leyland's connected-truck software iALERT, distinct from this vehicle manufacturing AI push, already rides in over 150,000 trucks across India.
Bharat Forge — Bharat Forge runs AI predictive maintenance manufacturing globally, fitting presses, furnaces and induction lines with IoT sensors and digital twins that flag faults automatically, cutting manual checks. A factory automation push adds robots like Snowy, a quadruped with cameras and LIDAR hunting oil leaks and joining robotic arms to fix misalignments.
Mankind Pharma — Custom GPTs built on OpenAI Enterprise now run throughout Mankind Pharma's workforce, spanning field force enablement, digital marketing, R&D, manufacturing and medical affairs. Teams handling sales and marketing have gained the most from this alliance with OpenAI, a case of AI in pharmaceutical operations reaching further than most rivals attempt. Hardly any pharmaceutical business in India has pushed generative AI so deeply into daily operations.
Pegatron — Pegatron's new AI-powered smart factory in Batam, Indonesia runs on a private 5G network and IoT sensors, launched under subsidiary PT Pegaunihan Technology Indonesia on 24 April 2025. Telkomsel equipped the plant with 1,200 SIM cards, letting machines report output and status in real time. The build, blending advanced automation with AI in manufacturing, supports a broader Industry 4.0 push and Pegatron's wider digital transformation goals.
Suzuki Motor Corporation — Suzuki runs AI work-analysis software called Ollo Factory, built by Tokyo startup Ollo, to monitor assembly-line tasks at its Sagara plant and, since December 2025, at its engine plant too. Smartphone, tablet or wearable-camera footage feeds the system, which breaks down worker motions, flags anomalies like a missed screw as they happen, spots inefficient movement, and helps train staff by contrasting beginners' actions with those of experienced workers using this AI-powered factory motion analysis. The deployment sits within Suzuki's wider Smart Factory programme.
Kansai Electric Power Company — Kansai Electric Power is deploying AI in utilities operations by rolling out ChatGPT Enterprise across its entire staff of roughly 8,300 employees, following a strategic partnership with OpenAI confirmed in June 2025. The utility fine-tuned the system using historical thermal-power-plant incident logs, applicable regulations and board-meeting records, producing a generative AI compliance tool capable of reviewing paperwork and responding to operational questions. A separate 'AI President' function, drawn from board records, lets employees query company strategy directly. The Yomiuri Shimbun noted plans to widen the deployment to group companies starting January 2026.
ArcelorMittal Luxembourg (Differdange site) — ArcelorMittal's Differdange plant now applies AI to scrap sourcing, furnace control and beam cutting, an AI in steel manufacturing rollout. Luxembourg reports lower CO2 emissions, less energy use, and new high-tech roles supporting the AI-powered steelmaking systems.
ILEVA — ILEVA's centre in southern and western La Réunion now runs an AI optical sorter, using cameras to pull recyclables from the waste stream. The €400 million Saint-Pierre facility housing it targets 240,000 tonnes diverted from landfill and 10% of island electricity within two years. This AI-driven waste sorting technology anchors a wider infrastructure push.
Nemak — Nemak, Mexico's car parts maker, rolls out machine learning manufacturing tech via NORIS at 35 sites globally, spanning its García, Nuevo León plant. Sensors log 500-plus metrics per unit, letting onboard AI trace faults' origins and forecast trouble before breakdown, a predictive maintenance approach catching snags early.
KIA México — AI factory inspection robot Spot, built by Boston Dynamics, patrols KIA México's Pesquería plant in Nuevo León unsupervised, starting each day at 6am with nobody present. Heat and compressed-air-leak sensors feed data back while it maps the facility into a digital twin, on a line producing 64 vehicles per hour. Two further Boston Dynamics machines, the Stretch logistics unit and Atlas humanoid, appeared on-site for autonomous robotic inspection manufacturing trials, still under evaluation.
AboitizPower — Digital twins, anomaly detection, early warnings and live asset checks now drive AI in coal power generation at AboitizPower's 300-megawatt Therma South Plant in Davao City, converted to a smart coal-fired station under Project Arkanghel. Next: the Therma Visayas Plant in Cebu and a Luzon site.
Korea Southern Power (KOSPO) — Korea Southern Power (KOSPO Evolving Mind Innovation) flipped the switch on generative AI public institution tool KEMI on 26 August 2025, marking what officials call South Korea's first such deployment inside a state-run body. Konan Technology co-engineered the platform, letting workers tap sensitive files - plant operations, energy policy, admin workflows - purely in-house, keeping outside leakage out. A companion avatar, Joy, guides consultations and delivers safety coaching, feeding a broader shift toward generative AI in the public sector.
Endress+Hauser — Endress+Hauser applies AI machine vision for assembly quality control at its Reinach plant, running MVTec MERLIC to steer manual builds and catch flaws instantly through semantic segmentation and object detection. Wired into the Manufacturing Execution System, this AI-assisted visual inspection has since spread to other plants globally, the Swiss firm says.
BMW Group — BMW Group's Debrecen factory in Hungary builds the electric BMW iX3 in series, with AI in car manufacturing woven through upkeep, parts flow and inspection. Following iFactory principles, the site pairs the AIQX quality platform with digital twins, self-steering carriers and a machine-run paint hall checked by AI-powered cameras. BMW ranks this AI-driven vehicle production hub its most innovative plant worldwide and first fossil-fuel-free vehicle factory, running wholly on renewables.
SSE Airtricity — SSE Airtricity's new AI energy advisor, the ESI Energy Advisor, now serves personalised savings tips to more than 65,000 smart meter households in Ireland. Built on Databricks infrastructure with Anthropic's Claude Sonnet 4.5 underneath, it mines billions of nightly meter readings to shape suggestions for individual homes. Within the advisor's first month live, 90% of logged-in customers clicked through to view their AI-powered energy savings recommendations, and unique visits to the insights section climbed 12% over the same period.
Zachry Construction trimmed 28 days from a $149 million Texas I-10 highway job using AI construction scheduling from ALICE Technologies. Modelling about 60 job orders took the AI scheduler around three minutes; the chosen plan shuffled tasks and reduced flat-work crews from nine to seven. Zachry and ALICE's founder claim the saving; no independent assessment has appeared.
Codelco — Iván Tobar, corporate maintenance manager at Codelco, Chile's state copper company, told a University of Santiago audience on 8 September AI in copper mining lifted production 40% and mineral recovery over 15% as technology use grew. The report does not identify the operations or period behind the figures, and no independent evaluation is published.
Cámara de Industrias de Costa Rica (CICR) — Costa Rica's manufacturers register almost no AI employment impact on staffing: 95.1% of businesses running artificial intelligence reported headcount unmoved, the Cámara de Industrias de Costa Rica found. Its Encuesta anual Perspectivas Empresariales y Valoración de Factores de Competitividad de la Industria Manufacturera, 2026 puts uptake at 53%, climbing past 71.4% inside free-trade zones. Hours, pay and task mix escaped measurement.
Codelco División El Teniente — Codelco's El Teniente division says AI workplace safety video analytics using 40 cameras has cut risky equipment–people interactions by 90% in monitored mine workshops. Specialists validate its automatic alerts; the system grew gradually from a workshop pilot in early 2024, and no independent evaluation is published.
Equinor — Norwegian firm Equinor's use of AI in oil and gas operations produced 1.3 billion kroner in combined savings and value creation during 2025, split with partners. Software watches 700 rotating machines for early faults, reads seismic data across 2 million square kilometres tenfold faster, and guided well planning that shaved 120 million kroner off one project — gains the company frames as AI in energy operations paying off.
AI could automate 39% of nonphysical work in construction and 50% in architecture and engineering, McKinsey estimates. This is automation potential, not current adoption.
National Grid Partners — 78% of utility innovation leaders now deploy or operationalise at least one AI application for interconnection demand, says the third annual Utility Innovation Survey from National Grid Partners. Another 74% blame AI data centre load growth for weaker grid reliability, and 84% need over a year from pilot to rollout. Method Research polled 134 leaders at utilities in the United States from 20 May to 7 July 2026, RepData distributing; thin AI skills and caution slow AI utility deployment.
Cámara de Industrias de Costa Rica (CICR) — The Costa Rican Chamber of Industries (CICR) finds, asking for the first time in its annual competitiveness survey, that 53% of manufacturers run some AI system or software. AI adoption in manufacturing hits 71.4% in free-trade zones versus 47.1% under the definitive regime, and 60.6% of large definitive-regime firms versus 38.9% of small and medium ones. Administration and business management (54.1%) and marketing or sales (52.5%) lead the uses, and 95.1% of AI users report no headcount change as a result.
ADNOC — ADNOC runs over 115 enterprise AI agents across auditing, procurement, finance and HR, chief executive Sultan Al Jaber told the UAE government's Agentic AI retreat. Some 20,000 staff trained to build job-specific AI models now use 3,000 daily, and company-wide AI utilisation reached 80% in the prior 90 days.
Regular AI use in construction rose to 19%, from just under 12% a year earlier, and the share of non-users fell by over a third, RICS's 2026 survey finds. It drew 1,883 construction and commercial-property responses in Q1.
86% of large US contractors expected an AI competitive advantage, against 69% of small and mid-sized firms, in a December 2025 Dodge Construction Network/CMiC survey of 235 contractors. Only 40% had a dedicated AI budget.
Parliament of The Bahamas — Bahamas' Electricity Act sets AI regulation in the electricity sector under Section 18, letting the Utilities Regulation and Competition Authority write binding rules for artificial intelligence use by power companies. Any such rules must satisfy ten named principles, among them safety, transparency, fairness, accountability and redress alongside security, robustness, explainability, governance and contestability. The statute defines artificial intelligence as tech letting a device or programme be trained or coded to sense surroundings, process data automatically, and produce decisions or suggestions — an unusually direct embedding of AI oversight in utilities law.
Sugarcane Industries Department, Government of Bihar — Sugarcane Industries Minister Sanjay Kumar launched Ganna Saathi, a National Informatics Centre AI chatbot for sugarcane farmers. Growers check permits, claims and applications; a digital farmer identity card comes too.
Consorzio Vino Chianti — Consorzio Vino Chianti has loosed AI for wine fraud detection on the open web, twinning artificial intelligence with tracking software to catch knock-offs and misuse of the Chianti name. Giovanni Busi's body disclosed on Wednesday 16 September that the net is cast over marketplaces, Facebook, Instagram, X, TikTok, sites, engines and domain registries. The ministerial scheme lasts to 20 January 2027, trademark enforcement AI.
JA Zennoh — CTE's NouNou Vision digital twin now runs full-scale for JA Zennoh in a Kanagawa greenhouse, bringing AI-based crop monitoring into daily use. Cameras feed images that software converts into 3D models, extracting stem length, leaf area and fruit tallies without manual checks, grounding greenhouse agriculture research in sensor data.
Development Bank Ghana — Development Bank Ghana (DBG) has rolled out Boa Mi, a WhatsApp-based AI-powered agricultural finance platform, alongside Opportunity International and the Ghana Incentive-Based Risk-Sharing System for Agricultural Lending (GIRSAL). Smallholders and extension officers get AI farm advice instantly; lenders see live portfolio data. A rice trial logged 8,500-plus queries over three months at 75 per cent satisfaction, with six more commodities next for this WhatsApp crop guidance.
Azerbaijan Food Safety Agency — The Azerbaijan Food Safety Agency will apply AI food safety risk analysis inside the Automated Food Safety Information System, or AQTİS, tightening enforcement, rule-making and the warnings that reach traders and shoppers, under a State Programme for 2027-2032 signed by President Ilham Aliyev. Kit selection comes first, jointly with other public bodies, before a model that flags hazard detection and brewing trouble is built and turned on. The Ministry of Digital Development and Transport and the Ministry of Economy share delivery.
UN Development Programme — A UN Development Programme pilot, backed by the Japanese government, blends satellite observation, field data and AI crop risk monitoring in Togo's Savanes region to catch crop threats before damage shows. Solafune, a Japanese satellite and geospatial data specialist, runs it over a 1,000 km2 pilot zone in the Cinkasse and Tone prefectures. In Lome, a Solafune team showed Togolese stakeholders early results and a test build of the satellite early-warning platform.
Ministry of Agriculture and Animal Resources — Rwanda's Ministry of Agriculture is trialling Tunga, an AI farm advisory voice assistant on the ministry's call centre taking farmers' questions in Kinyarwanda. It draws on a ministry, Rwanda Agriculture and Animal Resources Development Board and partner knowledge base, handing unresolved queries to a human expert. ICT and Innovation Minister Paula Ingabire says Tunga got 60% of test-phase questions right.
Sugar Cane Growers Fund (SCGF) — Fiji's Sugar Cane Growers Fund (SCGF) unveiled an AI chatbot for farmers, the Smart Chatbot, together with a new Digital Management System, moving the sugar sector in Lautoka away from paper-based work. Officials expect the digital farmer support tools to widen access to financing and information, particularly for women and other vulnerable groups. The Asian Development Bank backed the launch.
Ministry of Agriculture and Land Reclamation — Egypt has switched on AI in farm management: the Ministry of Agriculture and Land Reclamation now pairs satellite imagery with remote sensing to map farmland boundaries across 9.5 million-plus feddans and check soil and crop condition. Fertiliser flows through the 'Farmer Card' scheme; pests get caught early; harvests get forecast in advance. 'Hudhud', made with the Ministry of Communications, advises on watering and planting timed to climate shifts — one strand of AI-based agricultural monitoring spanning millions of feddans.
Ministério das Telecomunicações, Tecnologias de Informação e Comunicação Social (MINTTICS) through Gabinete de Gestão do Programa Espacial Nacional (GGPEN) — MINTTICS has rolled out AI agriculture monitoring at the Instituto Técnico Agrário do Quéssua in Malanje, using satellite imagery to track crop conditions and climate shifts through the Observa Mais platform. Built by the GGPEN space office, the system now reaches a third province after Luanda and Bengo, with fifteen students and one teacher trained to pass on the know-how. Uíge, Zaire, Lunda Norte, Lunda Sul and Cabinda are earmarked next, widening Angola's push for satellite-based farm and weather tracking.
Garden Grid — Mali's Garden Grid built Timbuktu, an AI crop monitoring tool pairing sensors and cameras to gauge soil temperature, humidity and light, spotting disease early and texting growers tailored advice. It also reads distant garden photos and links farmers to buyers via a paired B2B marketplace. The farm-monitoring technology is still trial-stage.
Confederação da Agricultura e Pecuária do Brasil (CNA) — Brazil's Confederation of Agriculture and Livestock (CNA) has launched JoIA, a free AI farm adviser on WhatsApp that technical director Bruno Lucchi says will serve as a technical consultant helping rural producers run their farms. Drawing on the CNA's Campo Futuro figures for input prices and production costs, the WhatsApp assistant lets producers compare themselves regionally and raise financing and debt questions, routing technical ones to Senar courses and training material.
Vanuatu Department of Forestry — The Pacific Community (SPC) and Vanuatu's Department of Forestry built a machine learning model for invasive plant mapping, reaching around 97 per cent validated accuracy for species like big lif rop, Cordia alliodora and African tulip. Officers fed it over 3,000 field points via Digital Earth Pacific, using predictive modelling for invasive species to guide site selection and treatment trials from August 2026.
Kakuzi Plc — Nut grading at Kakuzi Plc's macadamia processing plant now runs through an Intelligent Optical Sorting system, part of an AI optical sorting macadamia rollout that keeps output quality consistent without manual checks. Drones scan the Kenyan firm's fields for pest problems and crop condition, feeding analytics that shape when farmers plant and when they harvest. This AI-powered nut grading push sits within a broader shift toward machine learning and aerial monitoring as Kakuzi adapts to climate pressures on agriculture.
Kyrgyz Research Institute of Animal Husbandry and Pastures — The Korean International Cooperation Agency (KOICA) has installed 60 AI-enabled wearable collars on animals at the Kyrgyz Research Institute of Animal Husbandry and Pastures, marking a step into AI livestock monitoring for Kyrgyzstan. Bodit Inc. designed the underlying platform, Farmer's Hands, combining sensors and machine learning to catch early illness and behavioural shifts in herds. The intent is precision agriculture via continuous, automated animal welfare tracking.
National Agriculture and Forestry Research Institute — Laos' National Agriculture and Forestry Research Institute, partnering WHYBIZ Corp and KOTRA, trials the nation's inaugural AI livestock management proof of concept. RFID ear tags, remote servers and algorithmic surveillance show farmers each beast's identity, wellbeing, breeding and whereabouts instantly. First at the Livestock Research Center and a Brahman cattle farm, AI herd tracking should reach 18 further locations.
Federal Government of Nigeria — Nigeria's Federal Government unveiled NAPS, a national AI crop monitoring system built with OCP Africa and Ground Truth Analytics that reads satellite images refreshed every five days and pinpoints individual farms. Satellite farm mapping will spread in phases under a memorandum of understanding, from one state towards 15 priority states, following pilots the partners say reached 13 states and over 250,000 farmers.
Malawi University of Science and Technology — The Malawi University of Science and Technology (MUST) has introduced AI-powered fertigation, starting at a Thyolo campus greenhouse of tomatoes, bell peppers and cucumbers, under the TagDev 2.0 programme. Real-time crop growth readings trigger automatic irrigation and fertiliser dosing, continuously tuning water and nutrient levels. Automated fertigation should cut waste and raise efficiency.
Ministry of Agriculture of Zambia — Zambia's Ministry of Agriculture has switched on Mulimi360, an E-Extension Platform built around e-Mulangizi, an AI chatbot for farmers meant to replace scattered, face-to-face guidance with on-demand support. The rollout gives rural producers and extension officers one shared digital agricultural extension platform for crop advice, weather forecasts, early warnings, price trends and mapped land data. Authorities launched the sensitisation drive at the Mulungishi International Conference Centre, part of a broader effort to shift farming extension work online nationwide.
Cap Bourbon — OceanLive, an AI video analysis for toothfish system built by Thalos, now runs on Cap Bourbon's longliner Cap Kersaint in the French Southern Territories. Onboard sensors feed AI fish monitoring software that spots species and tallies légine australe as crews haul lines, tightening traceability. TAAF then picked Thalos to roll the setup out fleet-wide.
Ethiopian Institute of Agricultural Research — Early Newcastle warnings, enset bacterial wilt flags and chicken growth tracking now fall under one AI agriculture systems effort from the Ethiopian Institute of Agricultural Research (EIAR), built alongside the Ethiopian Artificial Intelligence Institute and GPS PLC. Dairy tools and poultry stress sensors join the trial at Holeta Agricultural Research Center, backing smarter livestock and crop disease detection nationwide.
Otsuka Foods Co., Ltd. — Otsuka Foods built Oishisa LENS, an AI flavour prediction system running at its Biwako Research Institute for Bon Curry, the firm's 58-year-old brand. The tool digitises 40 years of paper formulation records, letting the AI recipe prediction system forecast taste outcomes and suggest substitute formulas whenever ingredients change.
Ministry of Agriculture and Rural Development (Albania) — Albania's Ministry of Agriculture and Rural Development unveiled YLLI, an AI assistant for farmers, at a Tirana conference on digital agriculture through 2030. Housed in the Farmer's Portal, this virtual farming assistant answers in Albanian with guidance, market and collection-point prices and phone-based help; farmer Manjola Hafizi tried it live on stage, asking about the National Support Scheme 2026. Minister Andis Salla pointed to growing reliance on technology and real-time data across the sector.
Bakkafrost — GroAqua has delivered Bakkafrost's newest feed barge, Torvgjógv, now working the Fuglafjørður farm in the Faroe Islands. The 700-tonne vessel runs AI feed distribution aquaculture across 18 lines feeding 18 pens, the system now live and wringing more value from every load than older gear managed. GroAqua's Klaksvík feed centre tracks this AI-optimised salmon feeding remotely, expecting sharper efficiency, better oversight and a stronger environmental record for the sector.
Zyterra — Zyterra, a Zimbabwean agritech firm, built VerdantIQ, an AI farming platform pairing satellite crop data with weather forecasts and agronomic tips for smallholder farmers, aiming to sharpen everyday planting and harvest choices. WhatsApp and web access bring Mudhumeni Hungwe, a multilingual digital agronomist, straight to growers, widening reach for AI-driven agriculture support beyond costly apps. The firm aims to sign up 10,000 farmers free within 18 months.
Kamei Corporation — Kamei Corporation has begun trialling AI smart farming for asparagus at its ZAO AGRI BASE site in Miyagi Prefecture, where the Zero Agri system controls watering and feeding without manual input. Paired with raised-bed growing equipment, the setup forms part of a wider push into next-generation agriculture, letting Kamei track harvest size, crop quality and reduced labour needs. The trial marks an early test of this AI-driven agriculture system before any wider rollout.
Telenor Pakistan — Telenor Pakistan launched Kissan Dost Bashir via its Khushaal Watan agriculture platform, billing it as Pakistan's first conversational AI farm voice bot. Answering in Urdu around the clock via the 7272 helpline and 7272.pk portal, it offers live mandi rates, weather-based crop advice and livestock guidance to farmers facing literacy and connectivity barriers.
3Farmate — 3Farmate's FAMA, billed as Ghana's first AI-powered autonomous farming robot, was unveiled before Deputy Minister for Food and Agriculture John Dumelo. Computer vision replaces GPS for guidance, letting the machine plant seed, spread fertiliser, weed and spray while one worker oversees several units across 27 to 35 acres daily. Savings of 60% versus hand labour are expected, with design, engineering and assembly done locally. Officials pledged backing to expand this AI farming automation nationwide.
Econet Wireless Zimbabwe — An AI platform for farmers named EcoFarmer AI has gone live in Zimbabwe, built by Econet Wireless Zimbabwe as an artificial intelligence-powered agriculture platform to sharpen decision-making. It uses exact field coordinates to model conditions up to 100 days out, then suggests planting, watering and input timing from soil readings and local forecasts. The launch sits under Econet AI, the firm's AI arm set up April 2026.
Plzeňský Prazdroj Slovensko — Slovakia's last commercial hop yard, in Nemšová, is testing AI irrigation management for hop farming under Plzeňský Prazdroj Slovensko's PRE CHMEĽ project. Sensor stations record soil moisture, temperature, humidity and rainfall, feeding an app whose algorithm tells farmers when and how much to water. A prior Czech run of this smart irrigation approach produced a 40% output jump where staff followed its schedule.
First Mills — First Mills, a leading Saudi Arabian flour miller, has started running Kodamai's Kelvingrove AI agents in food production across its four facilities. The agent platform tracks procurement signals, supplier coordination and stock levels live, aiming for quicker decisions and clearer production visibility.
Council for Scientific and Industrial Research (CSIR-INSTI) — Snap a leaf, get an instant reading: CSIR-INSTI, the Council for Scientific and Industrial Research's Institute for Scientific and Technological Information, has built AI crop disease detection apps for Ghanaian growers. Its Maize Leaf Disease Detection App and Tomato Leaf Disease Detection App flag infections from one image, speeding up treatment. A companion Cropping Calendar tool adds locality-based sowing, watering, pest control and harvest advice, feeding wider AI-driven precision agriculture nationwide. Dr Paul Asante Danquah confirmed free, open online access.
Ibaraki Prefecture — Sagri has won a third consecutive contract from Ibaraki Prefecture to run AI crop surveys via its DETABA satellite imagery tool, verifying planting status for subsidy claims without manual field checks. Satellite data now replaces on-site inspectors, cutting officials' travel and labour costs. The renewal reflects rising confidence in satellite-based crop monitoring.
Saga Prefecture — Japan's Saga Prefecture named Sagri primary contractor for a demonstration project trialling AI field surveys to cut administrative spending on crop stabilisation measures. Sagri's software scans satellite imagery, flagging what is planted and which parcels need on-foot visits. Such AI-assisted farm checking could sharply trim paperwork.
Ministry of Digital Malaysia — Malaysia's Ministry of Digital, Bernas and the Ministry of Agriculture are taking Rakan Tani nationwide, targeting roughly 110,000 registered padi farmers. The co-pilot sends real-time, personalised AI farming advice over WhatsApp. Around 200 farmers were to receive AI agricultural guidance by end of March after a 10-week Perak pilot with 42 farmers.
Ministry of Agriculture, Livestock and Irrigation — A crop disease diagnosis AI tool called Farmer AI is now available from Myanmar's Ministry of Agriculture, Livestock and Irrigation, packed into a mobile app for growers nationwide. It also reads soil health and offers tips on sowing, reaping and selling produce, answering questions typed or spoken through its built-in chatbot. Speaking to officials in Nay Pyi Taw, Deputy Minister Dr Tin Htut noted the app remains Version 1 (Beta) ahead of a fuller Play Store launch, broadening access to AI-based plant disease detection.
Shaqodoon Organization — Somali NGO Shaqodoon Organization built SAADAAL, an AI early warning system giving farmers advance notice of droughts and floods. Partnered with Southwest State's Ministry of Humanitarian Affairs and Disaster Management, the tool reads live water sensor and weather station data, texting alerts straight to phones ahead of disaster. BBC News Somali covered this climate risk forecasting work for agriculture.
Ministry of Agriculture and Farmers' Welfare — Union Agriculture Minister Shivraj Singh Chouhan has launched Bharat-VISTAAR in Rajasthan, bringing real-time, personalised AI farm advisory to farmers. The multilingual platform had been announced in the Union Budget 2026-27.
CSIR-Crops Research Institute — Ghana's CSIR-Crops Research Institute (CSIR-CRI) runs SporeSmart, an AI crop disease prediction pilot catching airborne fungal spores and analysing images to flag outbreaks early. Five SmartSpore traps gathered over 2,000 spore photos between September 2025 and February 2026, targeting Fusarium pathogens harmful to maize, rice and tomato. UK Innovate funds the work with UK and Brazilian partners, advancing AI-driven disease outbreak detection.
Ethiopian Agricultural Transformation Institute — OpenAgriNet (OAN) went live across Ethiopia on 3 February 2026, unveiled by the Agricultural Transformation Institute (ATI) as a Digital Public Infrastructure serving as the countrywide backbone bringing AI in agriculture to Ethiopian smallholders. Farmers reach it mainly by voice, receiving guidance shaped by their own plot's crop mix, soil type and local climate. Open technical standards and consent-driven rules for handling personal data underpin the network, which ties together growers, extension officers, buyers, climate forecasts and lenders as one AI-powered digital platform for farmers.
Andorra Recerca + Innovació — Andorra Recerca + Innovació heads the pilot in BOSPIR-3D, an Interreg POCTEFA scheme trialling AI forest monitoring with LiDAR, drones and short-range scanners across high-mountain woodland. Teams opened work at Beixalís, La Rabassa and Arinsal following an April kick-off. Virtual reality joins artificial intelligence in reading and displaying the forest inventory data.
Ministry of Agriculture of the Republic of Azerbaijan — Azerbaijan's Ministry of Agriculture runs AI crop monitoring nationally, says Surkhay Novruzov, who heads its AI projects. Wheat stress readings began reaching the ministry in April and cotton has since joined, with every crop the target; a companion model judges whether water reaches tilled land, right roughly 82% of the time. A separate alert engine recognises 16 pests and diseases, pushing three-day outlooks to growers over Telegram at 07:00 daily, alongside evapotranspiration and plant-water figures.
Conservation International — An AI wildlife monitoring effort from the US-headquartered nonprofit Conservation International now picks up pileated-gibbon calls through listening devices at 10 spots in Cambodia's Cardamom Mountains, a south-western rainforest topping a million hectares. No camera reliably spotted the treetop-dwelling species, so researchers spent three months schooling a machine-learning species detection tool in the sound, tagging roughly half the nearly 800 calls in one six-week stretch. A parallel count of almost 150 units runs again this year.
Cayman Islands Department of Environment — The Cayman Islands Department of Environment now runs AI fish identification on the Nassau grouper aggregation off Little Cayman, shooting as many photographs as possible across three or four days for software that reads every marking. Croy McCoy, research manager at the Research and Assessment Unit, said postgraduates once tallied each grouper by eye on footage, hours of labour that facial recognition software absorbs while matching this year's arrivals against earlier ones. That evidence feeds the Grouper Moon Project, run with the Reef Environmental Education Foundation on an endangered species since 2001; a multi-agency crew returned in February 2026.
Kenya Space Agency and Ministry of Agriculture — Kenya Space Agency and Ministry of Agriculture launched AI crop monitoring nationwide through the Crop Measurement and Evaluation (CroME) Initiative at KALRO headquarters on 10-11 February 2026. With Microsoft AI for Good Lab and NASA Harvest, satellite-based crop surveillance delivers yield forecasting, crop maps, damage detection, field boundaries.
Mitsui Norin Co., Ltd. — Mitsui Norin's Sudama factory in Yamanashi Prefecture, Japan, runs AI OCR at full scale for raw material management, using Future Architect's Future EdgeAI. Staff snap packaging on portable devices to read lot numbers, product names and expiry dates. Character recognition makes receiving inspection through production line input paperless.
Gobernación de Antioquia — Antioquia's government will give 1,000 producers AI-powered agricultural assistance via its Agro100 pilot with the University of Antioquia and private-sector allies. Data capture, analytics and a chatbot-style helper will link farms to advice, coaching and online trading. Rural Colombia gains an entry into AI in agriculture.
Abu Dhabi Department of Energy and Abu Dhabi Agriculture and Food Safety Authority — Al Wathba's 80 farms have joined Abu Dhabi's Smart Meter Project, gaining AI-powered smart water meters wired into the Department of Energy's AD.WE platform. Growers now track usage instantly, tightening irrigation and cutting waste under the Agriculture and Food Safety Authority. Expansion follows to 130 farms in Al Nahda and 2,500 in Al Rahba, widening farm water management emirate-wide.
Philippine Rice Research Institute (DA-PhilRice) — Palai, DA-PhilRice's AI chatbot for rice farmers, reaches Filipino growers in local dialects around the clock via Facebook Messenger. It anchors the Philippines' first AI ecosystem for agricultural information, ending the Institute's 8 a.m.-to-5 p.m. weekday desk in favour of nonstop automated replies. PhilRice unveiled it Nov 25 at the 37th Ugnay Palay National Rice R4D Conference.
tacabAI — tacabAI now runs an AI-driven agriculture platform for smallholders across Somalia and East Africa, blending soil and crop sensing with photo-based disease diagnosis plus weather notifications. People can reach it by voice, text or picture and get answers in Somali, Swahili or English, easing access to AI farming advice.
Vacampo — Uruguay's Vacampo built CropIA, an AI precision agriculture tool overseeing 15,000 hectares. Satellites feed crop status, harvest trends and nutrient math into forecasts, giving growers an evidence-driven route to farm choices via AI in farming.
Agencia Espacial del Paraguay (AEP) — Small and medium growers in Paraguay gained a new AI in agriculture resource on 12 November 2025, when Paraguay Space Agency (AEP) unveiled Guaraní Insights: satellite readings on crop health and water stress from public imagery. Three phone-screen buttons unlock cloud-processed precision farming output built on a four-year archive, sharpening irrigation, pest alerts and harvest scheduling.
Ministry of Environment, Water and Agriculture — Under Saudi Arabia's Nama system, two AI agriculture platforms have gone live, backed by the Ministry of Environment, Water and Agriculture and already reaching over 1.6 million users. Slaughterhouses and broader protein supply chains sit under one tool's watch; crop demand projections for sustainable farming come from the other.
National Livestock Development Board — 480,000 coconut palms at Adigama farm, Dambadeniya, are now being scanned from the air under a National Livestock Development Board pilot, Sri Lanka's first attempt at AI coconut plantation management. Camera-equipped drones judge each tree's condition and flag what fertiliser or treatment it lacks, replacing manual checks. Namal Karunaratna, Deputy Minister of Agriculture and Livestock, cut the ribbon on the drone-based crop health monitoring venture.
Opportunity International — Opportunity International's Ulangizi delivers AI farming advice to Malawi's smallholder growers via WhatsApp, in Chichewa or English. This AI agricultural chatbot draws on official crop guidance and local know-how, accepting typed queries, voice clips and photos. It aids pest and disease diagnosis and helps growers cope with a shifting climate.
Federal Ministry of Agriculture and Food Security — Federal Ministry of Agriculture and Food Security has rolled out the Nigeria Agricultural Intelligence Platform (NAIP), an AI advisory tool built with Pixels Solutions for AI in agriculture. Farmers get weather outlooks, market guidance and agronomic tips, plus USSD access for rural areas without smartphones.
Ministerio de Agricultura y Ganadería (MAG) — El Salvador's Ministerio de Agricultura y Ganadería (MAG) runs AI crop monitoring through MIRA (Monitoreo Inteligente de la Red Agropecuaria), pairing Google satellite imagery with Google Cloud Platform processing to track crop health, flag droughts and floods, and estimate losses. Multispectral drones and IA models generate early warnings, directing response teams to the worst-hit zones. This farm monitoring system converts raw imagery into alerts.
Sang Hyang Seri — One worker now replaces a crew of ten, according to pilot Adam Dalah Agung, since Indonesia's state rice producer Sang Hyang Seri (SHS) began flying XAG P100 Pro units over 15 hectares of paddies. This effort, part of a wider shift toward AI-powered drones for rice farming, lets the machines handle spraying and fertiliser spreading without manual control, sharply cutting the labour once demanded by such tasks. SHS, which farms roughly 5,000 hectares altogether, wants to broaden its use of these autonomous agricultural drones going forward.
Petitti Family Farms — Petitti Family Farms, an Ohio horticulture business, now fields 25 Burro robots at Ridge Manor, up from 10 after adding 15 more units, early AI in horticulture via agricultural robotics. Each Burro Grande hauls up to 5,000 pounds, guided by vision systems and GPS for spacing, potting logistics and shipping. Unverified company figures credit the machines, running since earlier in 2025, with cuts to labour, fuel and upkeep costs.
MyBank and Ministry of Agriculture and Rural Affairs — MyBank (网商银行) and the Ministry of Agriculture and Rural Affairs' Information Centre put China's first AI farming assistant into service in 2025. It gives 24-hour agronomy, weather and finance support for over 200 crop and livestock types via MyBank丰收卡, grounded in regional soil and climate records from the Ministry. Tricky questions get routed to this AI agriculture advisory system's pool of over 1,000 verified specialists.
Super Krishak (Gham Power Nepal) — Gham Power Nepal has added a chat tool, Super Krishak Krishi Intelligence, to its Super Krishak platform, giving growers AI farm advisory in Nepali. Users ask about pests, weather, prices, fertiliser timing or cultivation and get answers suited to their own plot — AI-based agricultural guidance for local conditions.
Dialog Axiata — Farmers and home gardeners across Sri Lanka can now reach Govi Mithuru, Dialog Axiata's AI agriculture advisory service, through trilingual chat, voice interaction, photo-based pest detection, or a Sinhala-English call line at 616. All channels deliver expert-verified answers in Sinhala, Tamil and English, part of Dialog's latest step in AI-driven farming support.
JPI Farms — Zero round-the-clock monitoring, 22 hectares of peppers: JPI Farms, a Canadian greenhouse business, now runs Blue Radix's Crop Controller — an AI crop controller — full-time, having piloted it in 2023. Owner John Giesbrecht calls the autonomous growing tool a genuine "stress reliever". Cucumbers grow there too, bound for buyers in Canada and the United States.
Belize Forest Department — The Belize Forest Department and the Ministry of Sustainable Development launched three tools for AI forest monitoring, developed with the University of Alabama in Huntsville. Each analyses satellite imagery: land use cover map, near-real-time fire monitoring system and national mangrove cover map spanning 2020 to 2024. Chief Forest Officer Wilbur Sabido says AI-assisted forest tracking sits within the National Forest Monitoring System.
Agrosuper — Reaching over 55 thousand small shops, minimarkets and butcheries nationwide, MAI (Modelo de Atención Inteligente) gives Agrosuper's field reps a data-driven answer to an old logistics puzzle: whom to call on next. Chile's largest food producer built the AI in retail sales optimisation engine by cross-referencing billing volume, buying patterns and each outlet's location, then ranking visit priority and frequency for AI-driven sales targeting for small retailers. Coverage and service quality have both climbed since, though only company-reported figures back that claim.
Government of Kazakhstan — Kazakhstan plans a major expansion of AI satellite monitoring of farmland, President Kassym-Jomart Tokayev said on 8 September 2025. The upgraded system will assess soil quality, harvest volumes, and crop spread nationwide, with the same AI-driven land monitoring extended to forestry. A single digital land-resource map combining cadastral records is also planned.
Pupuk Kujang — A state-owned fertiliser producer in Indonesia, Pupuk Kujang Cikampek, has rolled out an AI farming assistant named Plantastic, an AI agricultural app reachable through WhatsApp to give growers instant guidance. Farmers can pose questions on fertiliser dosing and crop nutrition, or send pictures of ailing plants so the system can diagnose the problem and suggest a remedy. The company's own innovation team built the tool, unveiled at the Festinov 2025 conference.
Beef + Lamb New Zealand — Beef + Lamb New Zealand's new AI farming assistant, Bella, gives tailored guidance drawn solely from the B+LNZ Knowledge Hub: about 4,600 pages plus 220-plus hours of podcasts and video. Over two months, more than 100 testers, farmers, students and advisors among them, checked its answers before launch. Bella takes spoken questions in several languages, including Te Reo Māori and Filipino, functioning as a multilingual voice-based advisory tool for the sector.
Padiberas Nasional Berhad (BERNAS) — Malaysia's Padiberas Nasional Berhad (BERNAS) is piloting Project Rakan Tani, WhatsApp-based AI in paddy farming, within BERNAS's SMART Large-Scale Farming programme. Built with the National AI Office and Ministry of Agriculture, this AI farming assistant tailors hyper-personalised, real-time pest control, fertiliser and planting advice to each field, forecast and crop stage.
Wyke Farms — Wyke Farms, the UK's largest independent cheese producer, runs AI birdsong analysis across 10 South West England farms to gauge biodiversity. The three-year Birdsong for Biodiversity project with Chirrup.ai converts bird calls into species richness and ecosystem health readings, making Wyke Farms the first dairy processor using bioacoustic monitoring under the UK Dairy Roadmap.
Secretaría de Agricultura, Ganadería y Pesca — Argentina's Secretariat of Agriculture, Livestock and Fisheries has validated an AI method for soybean variety identification, using high-definition image processing on seed samples for real-time analysis. It determines predominant varieties and their proportions, reinforcing traceability and Seed Law enforcement. Seed variety classification starts at the Central Seed Analysis Laboratory.
Bayer México — Carlota, Bayer México's AI irrigation tool, sends WhatsApp irrigation advice to farmers in 16 Mexican states. This AI-powered irrigation advisory system uses satellite imagery, field sensors, predictive algorithms and big data; its 257-device network cut water use 30% on average, saving 14 million-plus cubic metres to date.
LAXEY — AI fish weight measurement replaces manual sample weighing at LAXEY's Icelandic sites, under a multi-year deal with Canadian firm ReelData for its ReelWeight cameras, an AI-powered camera system for salmon farming. Over 300 fish are scanned hourly, building on July 2025 trials. LAXEY will next pilot ReelAppetite, ReelData's AI feed-optimisation software.
Jets Sdn Bhd — Near the Muara port town in Brunei Darussalam, Jets Sdn Bhd has unveiled what it calls Southeast Asia's first net-zero-emission AI shrimp farm. Solar arrays supply energy while automated systems oversee blue shrimp breeding, blending renewable power with eco-friendly production methods under the AI in aquaculture approach. The firm describes the venture as a breakthrough for the nation's fisheries sector.
Connect Kisan — Farmers in Nepal can now speak their agricultural questions in Nepali to Connect Kisan's mobile app and hear spoken answers on topics from pest control to fertiliser choice, watering schedules and current market prices, an early step in AI voice advisory for Nepali farmers. Developed alongside Diyo AI, this AI-based farm advisory tool has already logged more than 2,000 queries in its opening phase.
Ariston Private Limited — Ariston Private Limited runs AI-powered sorting for macadamia nuts at its Chipinge estates. A camera photographs every kernel to spot flaws and rank quality, an AI-powered nut scanning method sorting produce by grade for buyers abroad.
Ministry of Agriculture of Jordan — Jordan's Ministry of Agriculture, with GrowTech Jordan, the Ministry of Digital Economy and Entrepreneurship, JICA and Japan Development Service, launched an AI soil fertility project pairing satellite imagery with machine learning to gauge land quality for agriculture, an AI in agriculture soil analysis approach. Farmers get data-backed guidance on irrigation and fertiliser use, cutting waste on both. The scheme runs through December 2025.
Federal Ministry of Communications, Innovation & Digital Economy — Over 1,000 farmers in Ogun State now carry smart devices with Huawei AI technology and MTN connectivity, issued by Nigeria's Federal Ministry of Communications, Innovation and Digital Economy under its AgriConnect Initiative. This early instance of AI in agriculture delivers planting advice, weather updates, market prices and financial data through AI-powered farming tools, sharpening decisions and cutting climate risk.
Alliance of Bioversity International and CIAT — Bruno, a camera-equipped robotic cart, gauges plant height, leaf spread and colour for the Artemis Project, rolled out by the Alliance of Bioversity International and CIAT to bring AI crop breeding to Uganda. Its scans feed the ONA mobile app, where algorithms swap in instant analysis for hands-on fieldwork. Aim: shrink a breeding timeline that normally runs five to seven years via this AI-driven plant breeding method.
Hyphen Tech — Hyphen Tech's AI crop disease detection tool AgriHyphen, trialled by Ngozi province growers, spots 37 ailments across seven staple crops from leaf photos, working in Kirundi, Swahili, French and English. Running since November 2024, the prototype earned praise for accurate readings, signalling agricultural AI adoption in Burundi.
Nitrofert — Nitrofert has released IngeDati, an AI in agriculture tool giving Colombian farmers crop-specific nutrition guidance across eight crops: coffee and rice, potato and banana, sugar cane, oil palm, citrus, plus pasture. A transformer-based architecture drives the system, built on recorded fieldwork and agronomic expertise, and farmers reach it via browser on ordinary phones rather than special equipment. Field tests under real growing conditions preceded launch, making this agricultural AI assistant a stand-in for in-person technical visits.
Agrix Tech — Agrix Tech's AI crop disease diagnosis app is giving Cameroonian farmers instant plant health readings, applying convolutional neural networks to photographed leaves to flag cassava mosaic virus, banana leaf spot and similar infections within 10 seconds at 95% accuracy. Offline functionality means the tool keeps working without connectivity, and it speaks French, English, Bamiléké, Hausa and further regional languages. Farmers trained by the startup reached 7,000 by April 2025 across the West and Northwest regions, with 5,000 users actively relying on this AI-driven plant disease detection tool; those users saw harvests rise 20-25% while cutting excess pesticide use by 30%.
Stift Klosterneuburg — Vitibot's electric Bakus now tends Stift Klosterneuburg's rows, Austria's debut in AI-assisted viticulture. Lighter and quieter than a tractor, the French-built machine loosens soil, mows grass, trims leaves and clears between vines after dark, easing compaction. A 75kWh charge keeps it running near ten hours, showcasing robotic vineyard management at one of the country's oldest estates.
Arla Foods — Arla Foods now runs AI milk supply forecasting across northern Europe, shrinking a multi-day task into hours while boosting precision by 1.4% for output tied to 1.5 million cows. Weather patterns, farmers shifting between milk categories, and each farm's daily volume feed the Danish cooperative's system. This dairy production prediction approach already spans every market Arla operates within Europe.
Gyeonggi Agricultural Technology Institute — South Korea gets its first AI smart farming pilot as Gyeonggi Agricultural Technology Institute pairs orbital imagery with machine analysis across 120 rice paddies growing the province's 'Younjin' variety and 50 fields under the Yeoncheon soybean research association. Sharpness is the selling point: satellites resolve ground detail down to 0.7 metres, four times finer than the 3m benchmark previously used, feeding growers near-instant readings on how crops are developing — a precision agriculture approach the institute hopes to expand. No results from the trial have surfaced yet, and the provincial body has not said when a wider rollout might follow.
National Electronics and Computer Technology Centre (NECTEC) — NECTEC has rolled out HandySense B-Farm, a smart farming platform pairing artificial intelligence with sensors and IoT technology to steer farm management. Live data feeds decisions on watering, fertiliser use and pest control, trimming costs and lifting yield quality. The launch backs a wider drive toward precision agriculture technology nationwide.
China-Pakistan Joint Laboratory for Agriculture — About 1,000 growers across Punjab and Islamabad have taken up Kisan 360, an AI-powered farming app from the China-Pakistan Joint Laboratory for Agriculture. Orbital imagery supplies moisture and nitrogen figures to users' phones, and a chatbot tuned on DeepSeek handles their questions, while drones fitted with visual-recognition tech spot pests and gauge plant stress. Both Apple and Android users can download the AI agriculture app.
Safaricom and Opportunity International — Safaricom and Opportunity International built FarmerAI, an AI chatbot for smallholder farmers in Kenya — an AI-driven farm advisory tool for potato growers, 800 to 1,000 of them, via WhatsApp on Safaricom's DigiFarm platform, covering weather, fertiliser, pests and prices. The trial spans February to August 2025.
Agro Prom — Armenia's Agro Prom runs an AI-powered greenhouse in Nor Yedesia where thousands of tomato plants are watched for disease, pests and ripeness with no human checks needed. Harvests already reach 18 to 20 tonnes a day, shipped onward to Latvia, Bulgaria, Romania and the UAE, and the firm hopes to push that total to 40 tonnes as the site expands. A patrolling machine hunts down insects, while an automated drip network handles all watering and feeding for the crop, an early example of AI in agriculture taking root locally.
Gouvernorat de La Manouba — Sensors now line artesian wells in Tunisia's El-Jedida delegation, feeding salinity, temperature and flow readings into an AI irrigation pilot run by La Manouba governorate. Algorithms cross-check this stream against satellite imagery to hand farmers exact watering guidance, cutting waste and lifting soil condition. PRIMA, the international funder, contributes 45,000 euros yearly out of a 2.6 million euro total, and a matching trial in Gabès governorate offers a comparison point for smart water management using AI.
Dundalk Bay Seafoods — Dundalk Bay Seafoods, a Co Louth family processor, runs Ireland's first AI-powered shellfish grader, scanning 10 seafood pieces per second via laser and machine vision. This AI shellfish grading system detects soft shells, cracks, sand and contaminants undamaged, starting with Dublin Bay Prawns and Razor Clams, backed by a €213,000 Seafood Development Programme grant.
JA Toyohashi (Toyohashi Agricultural Cooperative) — Aichi Prefecture's JA Toyohashi cooperative in Japan has scrapped pheromone traps for AI pest prediction in agriculture, running the TENRYO app built by Mirai Saien. Staff now flag outbreaks by blending weather data with farmer sightings instead of counting traps, cutting trap-based pest forecasting labour by about 90 per cent, down from over 40 hours monthly. No other JA cooperative nationwide has gone this far.
Royal Greenland — A computer vision AI for shrimp quality control, built by Royal Greenland with Trifork, cuts prawn inspection from 3-4 minutes to about 10 seconds per sample, at 95% accuracy on counting and recognition. Cameras judge colour and weight, replacing manual handling. Live at Cuxhaven, the AI-based prawn inspection system extends to Ilulissat, a Greenland site short of workers.
United Ajod Insurance — AI livestock insurance is now live in three Nepali districts, Humla, Lalitpur and Achham, run by United Ajod Insurance on muzzle-pattern biometrics from a smartphone app photo. That image becomes each animal's unique identifier, letting PlantSat's technology automatically match post-mortem pictures against the original record to settle claims. UNDP backs the digital livestock insurance rollout, which shortens payout times by making remote verification possible for herders.
Van Noord Growers — Eternal.ag's Harvester runs day and night for Van Noord Growers, a Zeeland tomato producer: autonomous tomato harvesting in greenhouse rows up to 22 hours daily, all week. Vision software judges ripeness per truss and cuts it free unaided. Trial success has the grower planning three robots for wider automated greenhouse harvesting.
Nelson Mandela African Institution of Science and Technology (NM-AIST) — KilimoAI, a smartphone app for AI crop disease detection built by NM-AIST's Dr Neema Mduma, spots pest damage and plant infections from a leaf photo. Over 22,000 Tanzanian farmers and extension workers use the tool, part of a farm pest identification effort targeting 400,000 users by 2030.
GanaderIA — GanaderIA's AI-based livestock weighing app pairs drone imagery with algorithms to tally herds and estimate weights, under 5% margin of error, skipping physical scales. Uruguay's platform converts footage into batch averages and per-animal figures automatically — drone-based cattle counting that replaces manual handling with remote data capture.
Rice Research Institute, Nukus / Uzbek agricultural authorities — A two-hectare AGRINOZE trial at Karakalpakstan's Rice Research Institute in Nukus cut water use by half without any fertiliser, holding yields steady. Buried sensors track soil conditions and pass readings to a machine-learning model, which times and doses drip-fed water and nutrients on its own — an AI-driven precision irrigation approach now piloted elsewhere in Uzbekistan. The result shows how an AI-powered irrigation system can shrink resource use while keeping output constant.
TechForward Tanzania — TechForward Tanzania rolled out Kilimotech Chatbot, an AI chatbot for farmers and livestock keepers, on WhatsApp. Users get immediate guidance in Swahili or English on financing, market prices, soil health, crop production and disease control, and the farming chatbot detects crop disease from a submitted photo.
PlantVillage Nuru, a free offline phone app for AI crop disease diagnosis from Penn State University, the FAO, IITA and CIMMYT, scored 65 per cent on cassava viral-disease symptoms in peer-reviewed trials, beating extension agents (40 to 58 per cent) and farmers (18 to 31 per cent). Its deep-learning plant diagnosis serves African smallholders and agents in the field and operates as a public good throughout sub-Saharan Africa.
University of Cambridge — Tessera, an open-source University of Cambridge model fed satellite imagery, got AI crop mapping right 84% of the time in Senegal's groundnut basin, Environmental Research: Food Systems reports. The satellite-based crop identification beat the next-best model by 28% in one test on far less labelled data and compute. The World Food Programme's country team weighs geospatial data and AI for food-security checks.
JA Zennoh — JA Zennoh trials put AI variable fertilisation rice gains at 5-15% above flat spreading rates. Z-GIS, the cooperative's farm platform, turns satellite NDVI scans into AI-generated fertilisation maps loaded onto Iseki and Yanmar transplanters, which throttle output automatically as crop vigour shifts within a plot.
Frontiers in Artificial Intelligence — Generative AI for agricultural extension went with measurably stronger advisory work among 240 extension agents sampled across six districts of northern Benin (β = 0.411, t = 8.229, p < 0.001). Agents carrying heavier caseloads and tighter deadlines picked up AI advisory tools more readily. Enthusiasm came hedged: respondents balked at leaning wholly on the technology, naming professional judgment, accountability and trust in AI outputs; the questionnaire was fielded March to May 2024.
Latakia University — Latakia University produced an AI fisheries stock assessment model inferring a Goldblotch Grouper's age and breeding status from total length alone, replacing monitoring war destroyed in Syria. Training used 150 Epinephelus costae from six coastal spots, March 2024 to March 2025, scales yielding age and first breeding at 30 cm. Scores hit R = 0.9995 on age and 100% on maturity, the largest specimen 74.3 cm and 12 years; the neural network stock tool suits unsurveyed waters.
Zoological Society of London — Zoological Society of London ecologists put AI seabird tracking at 93.65% accuracy for spotting red-footed booby dives from saltwater immersion records alone. Wet-dry logs came from lightweight leg tags; a parallel set of bulkier tri-axial accelerometers supplied the yardstick, and deep learning scored barely worse on the lighter tags. Machine learning dive detection could map feeding grounds all year from cheap sensors. Fieldwork ran at Barton Point, Diego Garcia, Chagos Archipelago, on a small pilot sample.
MHP — MHP, Ukraine's largest poultry producer, credits five years of AI in agribusiness with economic gains totalling millions of dollars. Taras Goshovsky put the 2025 figure alone at above $1 million, with combined earnings since the 2020 launch already well past that threshold. Sensors and predictive software now govern temperature inside poultry sheds moment by moment while guiding harvest projections across MHP's farmland, a use of machine learning in poultry farming that spans both livestock and crops.
NIBIO — Slashing herbicide volumes by up to 90 percent, without sacrificing cereal output: that is the result the Norwegian Institute of Bioeconomy Research reported from trials of DAT Ecopatch, a camera-driven tool for AI precision spraying. Its onboard vision distinguishes weeds from crop plants, releasing chemicals selectively instead of blanketing whole fields. Growers across Norway are already snapping up this weed-targeting spray technology.
In a four-year on-farm strip trial (2020–2023) on four heterogeneous arable fields totalling 49 hectares in Germany, sensor-based variable-rate nitrogen in winter wheat lifted overall nitrogen efficiency 15% and cut nitrogen fertiliser by up to 38 kg per hectare a year. The benefit was inconsistent: not every field beat uniform fertilisation.
Government of India (Ministry of Agriculture and Farmers’ Welfare) — Roughly half of surveyed growers in India shifted sowing plans once monsoon alerts covered 3.88 crore farmers across 13 states in kharif 2025. Phone surveys put the share revising land preparation, sowing timing, crop choice or input selection at 31 to 52 per cent, evidence AI in agriculture forecasting can shift smallholder behaviour at scale. Behind the alerts: an open blend of NeuralGCM, ECMWF's Artificial Intelligence Forecasting System and 125 years of IMD rainfall data, a farm-weather prediction approach using satellite and climate history.
In 2024 replicated trials of beet, spinach and pea at two US research stations, autonomous deep-learning laser weeding cut weed biomass by at least 97% against untreated controls and equalled or beat conventional herbicides. Laser-weeded peas out-yielded a bentazon plus imazethapyr programme by 15%. Manufacturer Carbon Robotics donated and ran the laser unit.
US farms using robotic 'box' milking earned dairy net returns 13% above non-adopters' in USDA's first nationally representative profitability estimate. Five Agricultural Resource Management Survey waves through 2021 feed the figure.
McKinsey & Company — Some 17% of the world's farmers use generative AI in farm-related tasks — among agriculture's fastest-growing technologies — says Global Farmer Insights 2026 from McKinsey & Company. Uptake runs fastest in the Americas, aiding daily decisions; robotics, electric-powered machinery and sustainability software barely register. Farm AI use was sampled from 5,500 farmers in 10 countries, April to July.
Ministry of Agriculture, Water and Irrigation — The Ministry of Agriculture's Malawi Digital Plant Health Services project reached about 157,000 unique farmer listeners with roughly 400,000 AI agricultural advisories by February 2026, via the PlantVillage app and Viamo voice service. Piloted since 2022, AI analyses field surveys for crop pests and diseases, reporting to a centralised AI observatory at Lilongwe headquarters. Lead farmers relay AI farm advice to smartphone-less neighbours.
In 2023 just 27% of U.S. farms and ranches used precision agriculture to manage crops or livestock. Even in a leading agricultural market, adoption remains a minority.
In 2023, 52% of midsize and 70% of large-scale U.S. crop-producing farms used tractor and harvester guidance and auto-steer systems. Early-2000s use was in single digits, and uptake rises steeply with farm size.
US precision-agriculture adoption climbs steeply with farm size. In 2023, small family farms, with gross cash farm income under $350,000, had the lowest adoption of every technology category measured, and retired operators and farms selling under $150,000 had the lowest of all.
Robotic milking has a measured profitability edge, yet robots milked the cows behind only 6% of US milk in 2021. Partly automated tools like computerised feed delivery covered about half of production.
Government of Namibia — Namibia's Cabinet told officials to scrap a N$40 million deal with US agritech company 6th Grain Corporation for countrywide AI crop monitoring by satellite. The Ministry of Agriculture, Fisheries and Land Reform then ended the Remote Sensing Agricultural Services Agreement, saying the satellite-monitoring deal broke the legal and procedural rules for state contracts. Earlier press coverage had queried how the tender was granted and whether an overseas company should oversee the nation's food production.
The EU's Digital Omnibus on AI, Regulation (EU) 2026/1744, took effect on 27 July 2026. It shifts the AI Act deadline for high-risk AI in Machinery Regulation products, such as harvesting robots and autonomous tractors, from August 2027 to 2 August 2028.
Sky Corporation — Sky Corporation's Gemini Enterprise deployment drove staff to build roughly 20,000 internal AI agents in five months across engineering, sales and administration, without central direction. Daily use of the Google platform doubled, with retention at 98%.
Bell Canada (Bell Cyber) — Bell Canada has put a cybersecurity AI model, built with Cohere, into production inside Bell Cyber's Autonomous Security Operations Centre to speed threat investigations. By condensing complex alerts and surfacing the context each case needs, the security language model trims analysts' information-gathering. Bell AI Fabric infrastructure hosts the system, keeping processing and sensitive data in a Canadian-hosted environment.
B3 — B3, Brazil's stock exchange, made Anysphere's Cursor its official AI coding assistant for engineering, widening it to about 600 employees after active developers grew from 34 to 154 between May and July. Luiza Sangalli, B3's superintendent of artificial intelligence, says application delivery time fell by about 35%, and the exchange now accepts roughly 1.5 million lines of AI-assisted code a month. Human review, testing and security checks still apply, and no independent measurement of the 35% figure has been published.
TechnoPro Holdings — TechnoPro Holdings has signed a company-wide licence for Anthropic's Claude Enterprise, extending AI in engineering services to over 30,000 staff. The deal folds the assistant into hiring, delivery and coding work group-wide, part of a wider AI workflow integration push. A 10 billion yen transformation budget also funds training 10,000 staff as 'AI implementation engineers' by June 2029.
NEC Corporation — NEC Corporation will roll out Claude to about 30,000 NEC Group employees worldwide, including developers and technical staff. The strategic collaboration expands AI in engineering and makes NEC Anthropic's first Japan-based global partner. An internal Centre of Excellence will develop AI-native engineering talent.
The US Federal Trade Commission issued 6(b) study orders, by a 3-0 vote on 11 September 2025, to OpenAI, Character Technologies, Meta, Instagram, Snap, xAI and Alphabet, examining possible harm to children and teenagers from AI 'companion' chatbots. It also checks children's online-privacy compliance, running alongside wrongful-death litigation against OpenAI (Raine v OpenAI, filed August 2025), a defining AI-safety case of 2025-26.
Ridery — Panita, Ridery's AI assistant on WhatsApp, racked up 830,000 monthly messages and 40,000 chat-booked journeys before its 2 December debut. Built with domestic talent, it organises taxi rides, film tickets, coach fares, cargo jobs and everyday queries in colloquial Venezuelan Spanish. Since 26 November 2025, it stands as the nation's pioneering WhatsApp virtual assistant.
University of Doha for Science and Technology (UDST) — University of Doha for Science and Technology now runs Microsoft 365 Copilot as a conversational front end for staff, part of a wider push into AI in university operations built on Azure AI Foundry. Agentic AI for higher education administration links departments, automating shared workflows including administrative processes. Qatar's MCIT backed the work, unveiled at Mobile World Congress 2025.
Siesa — Colombia's Siesa fields 49% of level-one technical support calls via an AI assistant on AWS. AI customer support here extends to engineering, where AI in customer support tools help staff write and test code; 20% of engineers are trained on them. Siesa's 10,000-plus clients span manufacturing, hospitality, retail and healthcare.
Nikita Mobile LLC — Arevik Vacharyan, an AI sales agent built by Nikita Mobile LLC, now runs outreach: spotting partners, negotiating terms, and passing strong leads to contracts staff without human input at each step. The Armenian firm reports solid early wins from this AI-driven sales prospecting abroad.
VSG Bulgaria — Every VSG developer now holds a GitHub Copilot licence, the Bulgarian firm's route into AI in software development for authoring code, generating tests and drafting docs. This AI-assisted coding push extends to analysts too, alongside a homegrown Retrieval-Augmented Generation setup for internal knowledge search and an automated tool matching insurance quotes.
Finnegans — Argentine IT firm Finnegans cut development times 29% with Finni, an in-house generative AI assistant built in under a month. Combining multiple language models, the workplace AI tool fields over 1,000 daily queries across all departments: customer service, HR, software development. Company figures since deployment show nearly 40% less non-technical work time.
NOVOMATIC Italia — NOVOMATIC Italia introduced EASY ASSIST, a WhatsApp AI chatbot for B2B customer support across AWP, Novocash and ADMIRAL Pay. The tool's dropdown menu handles 24/7 queries on change machines, spare parts and services. Built in-house in Rimini, Italy, the AI help desk should link to engines that learn from user interactions.
Grab — Grab debuted agentic AI for merchants and drivers across Southeast Asia at its first GrabX showcase. Its GrabMerchant app now hosts a chatbot, AI Merchant Assistant, suggesting tailored moves and automating ad campaigns or menu edits. An AI Driver Companion bundles demand-forecasting route tips with spoken hazard alerts, these AI-powered driver and merchant tools drawing 250,000-plus weekly users and 16,000-plus daily submissions.
Ciancoders — Quetzaltenango-based Ciancoders has released FinanzIA, marketed as Guatemala's first AI accounting platform built on automated bookkeeping technology. The software classifies documents, reconciles bank records, generates SAT-compliant ledgers and calculates shadow-declaration tax. Ciancoders claims an 80% cut in processing time and 15 new local tech jobs, unverified figures.
Wipro — Wipro has automated HR, finance and legal work with over 200 AI agents built alongside major cloud providers, running tasks independently rather than assisting staff, marking a shift in AI deployment in corporate IT services across India. Rishad Premji told shareholders about 87,000 employees have gained role-based training as agentic AI in enterprise operations becomes central to internal workflows and client offerings.
AI now writes about 75 per cent of Google's new code, which human engineers then review and approve, chief executive Sundar Pichai said at Google Cloud Next in April 2026, up from around 50 per cent the previous autumn. The on-record vendor claim rests on a loose definition of 'AI-generated'.
Spanish Data Protection Agency (AEPD) — Spain's AEPD, the Spanish Data Protection Agency, has logged a first AI agent data breach: an autonomous attacker built on a widely known language model altered personal records and opened invoices. After probing generic files and signing in, the agent explored the application unaided until exploitable flaws surfaced. Neither victim nor model is identified, and details rest solely on the affected organisation's report, pending analysis of this agentic AI cyberattack.
About 44% of AI code-generation tasks added a security vulnerability in Veracode's 2026 GenAI Code Security Report, whose average security 'pass rate', about 56%, is essentially flat on last year's 55%. AI-written code is no more secure even as its volume surges.
AI adoption follows a 'J-curve', productivity first dipping during rollout and then rising, Google Cloud's 2026 DORA research ('The ROI of AI-Assisted Software Development') finds. It confirms AI magnifies whatever strengths and weaknesses an organisation already has.
METR's randomised trial of 16 experienced open-source developers (Cursor Pro, Claude 3.5/3.7; Feb-June 2025) showed tasks taking about 19% longer with those early-2025 AI tools, although participants predicted gains and still felt quicker afterwards, so self-reported AI coding productivity cannot be trusted. METR's follow-up cohort (from Aug 2025) and 24 Feb 2026 update point to probable gains with today's tools, lower bounds of about 18% for returning developers and 4% for new recruits, with developers likely quicker in early 2026 than early 2025. Because developers who value AI most increasingly refuse to work without it, selection effects make task-level developer productivity hard to gauge, and METR is reworking its study design. Another METR survey of 349 technical workers (Feb-Apr 2026) recorded median self-reported gains of 1.4-2x in work value and 3x in speed, which METR warns are self-reports rather than measured effects; its own staff gave the smallest figures.
Developer trust in AI is falling as adoption rises. In 2025, 46% of developers actively distrusted AI tools' accuracy against 33% who trusted it, and favourable sentiment fell to 60%, from over 70% in 2023 and 2024.
Infosystems Jet and Smart Ranking — 86% of large Russian companies now use or pilot large language models, close to McKinsey's global rate of 88%, per research unveiled at the State of AI Russia 2026 event by Infosystems Jet and Smart Ranking. This LLM adoption in Russian companies drops off sharply at higher autonomy: semi-autonomous agents run at only 15% of firms, and fully autonomous systems at just 8%. Respondents named GPU expense (58%) and legal exposure (56%) as the leading obstacles to wider enterprise AI adoption.
Equipos Consultores — Equipos Consultores finds half of Uruguay's adults have never used AI tools; a quarter are daily or fairly frequent users and the rest occasionally. AI use in Uruguay tracks education: almost nine in ten university graduates have tried it, but eight in ten people with only primary schooling never have. The survey of 705 adults on AI habits ran from 23 June to 8 July 2026.
DOBA Business School — Slovenia tops the AI maturity index among Slovenia, Croatia and Serbia, scoring 50.85 out of 100 in DOBA's DAIMI 2026 survey - nearest of the three to the 'integrated' benchmark. Firms were surveyed between March and May 2026 across five dimensions: strategy, governance, workforce skills, deployment and business impact. Even progress on all fronts drives this AI adoption readiness lead, rather than strength in one area.
Moldova Innovation Technology Park (MITP) — A new poll shows 78% of firms at Moldova Innovation Technology Park (MITP) have woven artificial intelligence into daily operations, reflecting broad AI adoption in Moldova IT Park. Researchers surveyed nearly 2,800 businesses between April and May 2026: 37% run it organisation-wide, 41% apply it to select tasks, 10% are trialling it, and 11% reported none. Findings point to workplace AI tools spreading across Moldova's tech sector.
HKIRC — A poll of 800 organisations by Hong Kong Internet Registration Corporation found 94% had adopted AI tools at work, reflecting widespread AI adoption in Hong Kong businesses. Yet 63% had no formal policy covering workplace AI tools and 68% skipped staff training on the technology. Unauthorised AI tools were in use by staff at nearly half the firms surveyed, without company approval.
1,154 respondents across 32 provinces went into a national survey of AI use in Afghanistan, finding 88% of young Afghans familiar with artificial intelligence and 66.6% of them regular users. Media outlet Salam Afghanistan and Salam Watandar Radio's eight-month fieldwork, split 50.6% men and 49.4% women, found 84.1% had sought AI advice at least once and 82% now lean less on friends and relatives, with 58.1% uneasy about machines making decisions.
Eurostat — North Macedonia trails the EU in AI tools adoption among young people, Eurostat's 2025 data shows: 45.4% of those aged 16-24 use such tools, against 63.8% across the bloc. Generative AI usage among all residents aged 16-74 reached just 22% that year, leaving a wide gap versus the EU average for both youth and the wider population.
Pigé survey for Le Quotidien — AI tools Réunion adoption stands at 73% of internet users, a survey by Pigé for Le Quotidien found, matching levels seen in mainland France. The poll of 663 respondents aged 15 and over found artificial intelligence use has overtaken e-commerce in uptake speed, despite arriving on the island only in late 2022. Convenience and performance ratings for AI beat search engines, though trust and reliability lagged by up to 25 points.
AIPA (Yapay Zeka Politikaları Derneği) — Corporate AI adoption in Turkish business reached 39.2% this year, up sharply from 14.5% in 2021, according to a joint poll of 992 companies across 12 provinces by the Artificial Intelligence Policies Association (AIPA) and Kuantum Araştırma. Marketing topped the list of departments putting the technology to work, at 42.9%, ahead of management, finance, R&D, HR, production, public relations and legal — a sign that enterprise AI adoption in Türkiye is spreading well past its initial marketing foothold.
Centraal Bureau voor de Statistiek (CBS) — AI adoption in Dutch companies has doubled since 2023, with Statistics Netherlands (CBS) counting one in six firms as adopters — a trend also described as growing business use of artificial intelligence in the Netherlands. Mid-sized organisations employing 50 to 250 people saw the sharpest gains, up from 20% to 45%. Sales and marketing topped applications among adopters at 35%, ahead of administration at 32% and R&D at 25%.
Amazon Web Services — Nearly 170,000 Singapore businesses, 48 per cent, had adopted AI by November 2025, versus 40 per cent a year earlier, an AWS-commissioned survey of 1,000 business leaders shows. Financial services leads business AI adoption at 71 per cent, then technology at 70 per cent and healthcare at 63 per cent. Of firms with AI uptake, 82 per cent reported revenue gains averaging 19 per cent.
SAP — SAP-commissioned research by YouGov finds 81 percent of enterprises in Saudi Arabia already use industry-specific AI, with results unveiled at SAP NOW AI Tour Riyadh. Looking ahead, 96 percent intend to fund data consolidation programmes. Some 52 percent of respondents judge sector-tailored AI's success by customer satisfaction gains.
Statistics Norway (SSB) — Generative AI adoption Norway: Statistics Norway puts the figure at 54 percent among Norwegians aged 16 to 79 in 2025, up from 36 percent in 2024. Businesses with ten or more staff show similar growth, with 30 percent using AI technologies, three times the 2023 rate. Firms without such tools mostly blame a shortage of relevant skills, cited by 77 percent.
Manx Telecom — 52% of managers on the Isle of Man now report using AI at work, according to Manx Telecom's first island-wide AI adoption survey, run by Island Global Research among 640 residents. Across all employed respondents the figure drops to 43%, with heaviest use in IT operations (50%), marketing and sales (46%), and R&D (42%). Two-thirds of users, 65%, say productivity has risen as a result, despite only 15% ever receiving training on the technology, an artificial intelligence uptake gap the study flags as a risk.
In 2025, 84% of developers used or planned to use AI tools in development, up from 76% in 2024. 51% of professional developers now use them daily.
AI adoption went with higher software delivery throughput but, again, lower stability in Google Cloud's 2025 DORA report. It surveyed nearly 5,000 technology professionals, 90% of whom use AI at work.
Amazon Web Services — AI adoption in Malaysia has climbed to 2.4 million firms, 27% of the total, per a commissioned Amazon Web Services poll of 1,000 companies, a 35% rise from 20% the year before and 630,000 newcomers to the technology. Still, most stay shallow: 73% lean on chatbots and ready-made software rather than custom builds. Smaller players outpace larger ones among Malaysian businesses using AI, with 48% weaving it through daily operations and 31% crafting AI-driven products, versus only 44% and 15% at bigger firms.
US Federal Trade Commission — The Federal Trade Commission is examining AI safety broadly, covering Anthropic and OpenAI, a senior FTC official confirmed. This US AI regulation step tests claims of unfair or deceptive conduct and possible consumer harm; formal information demands may follow under federal AI oversight.
Office of the Governor of New York — Governor Kathy Hochul has timetabled New York AI regulation under the RAISE Act: large frontier AI developers register from November and prepare transparency, safety and incident-reporting processes, flagging safety incidents within 72 hours. Full compliance with the frontier-model rules begins January 2027, reporting to DIGIT, the Office of Digital Innovation, Governance, Integrity and Trust in the New York State Department of Financial Services.
Office of the Governor of California — Governor Gavin Newsom's executive order on California AI regulation fast-tracks third-party safety and security checks on AI systems, plus independent audits, under the new state law that subjects AI companies to outside oversight. Experts get two months to draft a guide to hardening the state's AI safety oversight, and the order advances an "AI kill switch" for frontier models.
Office of the Governor of California — Governor Gavin Newsom signed SB 1119, AI companion chatbot safety rules mandating suicide-crisis protocols, parental controls and alerts if a child disables safety settings. The chatbot protections for minors also make companies commission independent child safety audits and yearly risk assessments, a national first.
State of California — California Governor Gavin Newsom signed two bills his office calls first-in-the-nation standards for independent assessment of AI systems. Senate Bill 813 frames independent verification organisations that check AI systems and models for state-law compliance; Assembly Bill 1405 creates a state registry of AI auditors with independence, transparency and integrity standards.
Government of Japan — Japan's government has adopted principles on AI training data disclosure, pressing generative AI operators to share basic facts with users, rights holders and the public. Operators should publish an overview of their automated online data-collection tools, tell rights holders on request whether their works were used, and file their implementation with the Cabinet Office, updating it yearly. The non-binding principles carry no penalties and cover foreign as well as Japanese firms offering AI services in the country.
Colorado Department of Law — The Colorado Department of Law published draft rules on 11 August for the state's Automated Decision-Making Technology Act and its conversational AI service provisions, setting compliance duties for developers and deployers. Signed by Governor Jared Polis on 14 May as SB 26-189, the ADMT Act repealed and replaced the 2024 Colorado AI Act with a narrower framework on transparency and disclosure of adverse outcomes from automated decisions, dropping mandatory risk-management programmes, impact assessments and the duty of reasonable care. Comments are invited before a 26 October public hearing, and the Act applies from 1 January 2027 whether or not rulemaking is finished.
State of California — California's AI Transparency Act took effect on 2 August, obliging publicly available generative AI services with over one million monthly users in California to add latent disclosures to AI-generated audio, video and images, offer an optional manifest disclosure and publish a free AI detection tool. Passed as SB 942 in 2024 and amended by AB 853 in 2025, it carries civil penalties of $5,000 per violation, each day counting separately, with no private right of action. Hosting and large online platforms face further duties from 1 January 2027, and capture-device makers from 2028.
National Assembly of Vietnam — Vietnam's first standalone AI law took effect on 1 March 2026, enacted on 10 December 2025, grading AI systems high, medium or low risk by effect on safety and human rights, field of application, user scope and scale of effect. Providers must tag AI-generated audio, image and video in machine-readable form, and deployers must flag it where it might blur whether events or people are real. Baker McKenzie says most existing providers and deployers have until 1 March 2027, and the Ministry of Science and Technology will issue implementing guidance.
Infocomm Media Development Authority — Singapore launched the Model AI Governance Framework for Agentic AI at the World Economic Forum in Davos, covering AI agents that reason and act autonomously for users. The Infocomm Media Development Authority sets out technical and non-technical measures for agentic AI governance across four dimensions: bounding risks up front, meaningful human accountability through checkpoints, technical controls over the agent's lifecycle and end-user education. Minister for digital development and information Josephine Teo announced it.
Instituto Nacional de Calidad (Inacal) — NTP-ISO/IEC 42001:2025 gives Peru its first technical standard for AI management systems, set by the National Institute of Quality (Inacal) under the Ministry of Production. Based on ISO/IEC 42001, it guides organisations in building and improving such a system, covering oversight, targets, hazard evaluation and impact review — Peru's new yardstick for artificial intelligence governance.
Codelco — Chile's Codelco opened 5,500 Microsoft 365 Copilot licences for generative AI at work across divisions and head office, via an MoU with Microsoft. Training, change management, security safeguards and 600 'digital promoters', chosen from 900 sign-ups in three days, back the AI productivity rollout; agents aid procurement, legal and sustainability.
Assaí Atacadista — Assaí's internal auditors now use Auditus, a generative AI audit agent Vivo assembled on Microsoft Copilot Studio following an isolated proof of concept. Auditus interprets documents, checks replies against company policy and proposes technical lines of enquiry inside Teams, SharePoint, Word and Copilot Chat. Vivo pitches the co-pilot for auditors as backing judgement calls and trimming routine analysis.
Cebu Pacific — Cebu Pacific trimmed mean legal contract review from 2.5 hours to roughly 1.2 hours on ChatGPT Enterprise, working alongside Thinking Machines Data Science and OpenAI. Beyond AI contract review, engineering lookups fell from 10 to 15 minutes to beneath 10 seconds and project triage from five days to about one, shedding 80% of hand checking. 86% of polled staff said automated contract checks and peers carry over a third of daily duties.
Bukhatir Group — Bukhatir Group has switched on DIWAN, an enterprise AI assistant, for staff at all 16 firms it operates out of Sharjah. Magure, an enterprise artificial intelligence house in the UAE, wrote DIWAN on its MagOneAI agentic platform. The private artificial intelligence deployment pins the model to Bukhatir's own tenancy, so prompts and files never travel beyond the group's servers, under group policy and with every use logged.
Cisco — Cisco has given its whole workforce of about 90,000 people MyAgent, a personal AI agent built on its Circuit AI platform. Operations executive vice-president Thimaya Subaiya said it takes an objective and coordinates the steps, running autonomous workflows across apps such as Outlook, Webex, Jira and SharePoint.
Sportfive — Sportfive has put generative AI in sports marketing work, giving Microsoft 365 Copilot to all 1,900 staff in 20 countries with 93% adoption and 88% satisfaction reported. Employees use it for translation, document creation, outreach drafting and sponsor research, after everyone passed 'AI driving licence' training aligned with the EU AI Act.
Azm Saudi Arabia for Communications and Information Technology — Azm Saudi Arabia for Communications and Information Technology has five AI employees inside operations: an assistant to the chief executive, a business analysis employee, a talent recruitment employee, an IT services and facilities employee, and a quality and monitoring officer that tracks how each technical project is built and raises faults early. Each carries an identity, works over email and other internal channels, and meets the standards set for human staff. Azm puts the lift in operational efficiency above 30 percent, its own figure for agentic AI in enterprise operations.
Mews — Hospitality software firm Mews has deployed Duna's AI compliance agents across customer onboarding, from sanctions and PEP screening to know-your-customer checks and ongoing monitoring. Chief executive Matthijs Welle says the agents cut manual reviews and false positives and raise straight-through processing.
NEC — NEC has set up an AI agent department with no human staff: 'worker' agents run automation tasks under AI 'manager' agents, which answer to human supervisors. The company aims to lift productivity by separating human and AI work more clearly; the department's size and output were not disclosed.
Abu Dhabi Government, Department of Government Enablement — Abu Dhabi Government is rolling out Microsoft 365 Copilot for civil servants: its Frontier Employee Programme added 26,000 users across 27 government entities to 9,000 existing licences, making 35,000. Wesam Lootah, Director General of GovDigital at the Department of Government Enablement, says the public-sector AI rollout stays governed and secure, with Advanced Data Residency on every licence keeping processing inside the UAE. AI training and certification come with it.
Nutun — More than R40 million in outstanding balances across over 80,000 accounts has already been cleared by Zoey, the automated voice system built by Nutun for AI debt collection. Unveiled alongside the firm's new AI Operations Centre in Sandton, Johannesburg, the locally developed assistant now handles repetitive recovery calls on its own. Staff have shifted into roles overseeing and reinforcing the technology, an AI voice agent for debt recovery, instead of losing their jobs to it.
AlRayan Bank — AlRayan Bank adopts Microsoft 365 Copilot for bank staff under the Qatar Central Bank's AI guidelines, covering presentations, data analysis, document drafting and summarising reports and emails, it announced on 20 May 2026. Teams are trialling the AI assistant for credit documentation automation and analysis, ESG analysis and ratings support, and mobile-app user stories.
Government Commissioner for Artificial Intelligence — Palkovics László, Hungary's government commissioner for artificial intelligence, has unveiled AI Start 500, handing an AI management system for SMEs to 500 Hungarian small and medium-sized enterprises. Plugging them into AI-based business processes is the purpose, and the commissioner cast the scheme as a small-business AI platform that dissolves firms' wariness.
Tribunal de Contas do Estado do Paraná (TCE-PR) — Paraná's state audit court, the TCE-PR, has rolled Microsoft 365 Copilot into its work at a staff launch, putting AI in public auditing alongside administrative and technical tasks. A training track follows, with basic AI literacy compulsory for every employee.
Investec — Investec has given all 8,000 employees Microsoft Copilot across South African, UK and wider overseas businesses, calling it the country's first whole-workforce Copilot launch. More than 800 AI agents in banking now handle everyday analysis, drafting, reconciliation and admin work group-wide, saving over 350,000 staff hours annually by Investec's estimate. Lyndon Subroyen, global head of digital and technology, backs the bank-wide AI programme with training and an AI governance framework.
SOMPO Holdings — SOMPO Holdings began rolling out an in-house AI agent for employees, about 30,000 across domestic group companies, from January 2026, including all staff at SOMPO Holdings and core operating company Sompo Japan. With Google Cloud's Gemini Enterprise as the main tool, the group is testing what the agent actually changes about work. The goal is a foundation for reworking practices and processes rather than tinkering with them.
CCI Mayotte — CCI Mayotte now offers AI business creation support via IDcreaCCIon, a digital platform from the Chamber of Commerce and Industry of Belfort steering founders from concept to launch. Advisors pair the software's output with their own know-how, covering funding, legal structure, financial forecasting and market study. Mayotte is the tool's first outpost beyond mainland France, already relied on by several French chambers for entrepreneurship support.
TP Indonesia — TP Indonesia (formerly Teleperformance) unveils TP.ai FAB for AI in business operations, fusing generative AI, predictive modelling and analytics under one roof. Live sentiment reads and next-best-action cues steer telesales and collections, while automation absorbs routine tasks. This AI-driven business services push centres people, augmenting rather than substituting them.
Ooredoo Qatar — Ooredoo Qatar has broadened access to Microsoft 365 Copilot, rolling out the licence to a larger share of its workforce under an agentic framework for AI-driven workflows across the company. This Microsoft 365 Copilot enterprise AI rollout is meant to lift productivity and spur innovation through autonomous, AI-powered processes throughout the organisation. The move reinforces Ooredoo's standing as an early adopter of enterprise AI deployment in Qatar's telecommunications sector, extending the licence beyond its initial user group.
KPMG LLP — KPMG's whole US staff can tap Gemini Enterprise, Google Cloud's agentic AI platform, spanning audit, tax and advisory work. Roughly 90% signed on within a fortnight, churning out almost 700 no-code AI agents since late September 2025. Client-facing versions serve finance, procurement and infrastructure needs, part of growing AI-driven professional services momentum.
Mitsui Fudosan Co., Ltd. — Japan's Mitsui Fudosan wants AI in real estate to trim firm-wide hours by over 10 per cent after handing roughly 2,000 employees ChatGPT Enterprise. Workers built about 500 Custom GPTs in a quarter, guided by 150 AI Promotion Leaders in 85 departments. Bespoke property AI includes President AI Agent, mimicking President Takashi Ueda's judgement, and Automated Document Generation AI.
Severstal — Severstal, the Russian steelmaker, now offers Da Vinci, letting staff build their own workplace AI assistants, marking wider AI in steel manufacturing. No coding is needed; it handles text and images, reviews documents, and links to company systems. About 500 workers run roughly 200 setups, from ticket handling to meeting summaries to vibration checks. This generative AI platform for employees goes company-wide in 2026 with multimodal and autonomous agent features.
Altipal — A 45% cut to operating costs followed Altipal's launch of ORUS-brAInChat, an AI ecosystem for retail distribution built on the Model Context Protocol (MCP). Query turnaround at the Colombian distributor fell from 4.2 hours to minutes, this back-office AI tool automating finance, logistics, HR and project teams. Uptake hit 70%, lifting its Net Promoter Score to 78.
Zangia.mn — An Advanced AI engine now runs inside Zangia.mn, scoring CVs, pairing applicants with openings, and flagging roles by skill set and location. AI recruitment in Mongolia advances as career analysis and auto-written CV summaries reach 1.2 million users and 10,000-plus firms on the site, part of a shift toward AI-driven hiring software.
Poste Italiane — Poste Italiane rolled out postemyvoice, letting BancoPosta and PostePay customers in Italy prove identity by phone through speech, with no passwords or private information. Callers enrol once via voicebot, and voice biometric recognition then spots them on every later call. Nuance from Microsoft powers the voiceprint matching.
AIB — Microsoft Copilot banking now touches 10,000 AIB staff via Microsoft 365 apps — Outlook, Word, Excel, Teams, PowerPoint — cutting time spent on routine admin. AIB's AI Centre of Excellence is also crafting bespoke tools with Copilot Studio, extending AI in banking beyond a single off-the-shelf tool.
e& UAE — e& UAE has embraced AI in telecom operations, pushing Microsoft 365 Copilot out to staff company-wide. Coding work for hundreds of its engineers now runs through GitHub Copilot, while Azure OpenAI sits inside internal apps organisation-wide. The initiative reflects the operator's ambition toward an AI-powered workplace.
Korean Institute of Certified Public Accountants (KICPA) — South Korea's Korean Institute of Certified Public Accountants (KICPA) unveiled ChatCPA, a beta member tool delivering AI accounting and tax help via a large language model. Cited answers pull from Korean International Financial Reporting Standards (K-IFRS), prior inquiry records, plus relevant tax rulings and precedents. Members gain an AI-assisted tax research helper with built-in source traceability.
DBS Bank says its AI value in banking reached about S$1 billion in 2025 from more than 2,000 AI models across 430-plus use cases, up from over S$750 million across 370-plus in 2024. The figure, from the annual report's chairman and CEO letter, is DBS's internal estimate of costs saved and revenue gained, unaudited, and harder to isolate as AI embeds in daily workflows.
Proximus — Proximus runs its AI Energy Box system across 44 sites in Belgium, trimming power draw and lifting renewable output through AI-powered energy management. Companion.energy, a Belgian startup, paired a digital twin with live market feeds so signals steer cooling gear and batteries toward greener consumption. A 2024 trial came first.
3HORIZONS / CIO Vietnam — Just 15% of firms see AI business value as higher sales, despite roughly 95% using generative tools individually, finds a 3HORIZONS study with CIO Vietnam. Enterprise AI returns trail uptake: 59% use it in workflows, 36% in applications or processes, about 7% in autonomous systems; 44% trial agents.
Federal Reserve Bank of Philadelphia — AI adoption by small businesses reached 37% in the Third District (Delaware, New Jersey, Pennsylvania) during autumn 2025, versus 46% nationally, per the Philadelphia Fed's reading of the Federal Reserve's 2025 Small Business Credit Survey. Writing and marketing (79%), notetaking and summarising (56%) and research or planning (51%) led small-firm AI use. Productivity rose for 71%, yet over 70% reported labour costs, outside services, quality and sales unchanged.
BambooHR — US workers spend more AI time troubleshooting errors and reworking prompts (42%) than on productive work that advances their workload (35%), a BambooHR survey of 1,600-plus full-time salaried workers finds. Respondents used AI an average 87 minutes a day, and BambooHR estimates roughly 20 workdays yearly go to troubleshooting, though 65% felt confident and enthusiastic about AI at work.
Federal Reserve Bank of New York — AI job cuts remain rare in the New York and Northern New Jersey region: only 4 percent of service firms and no manufacturers shed staff over AI, a New York Fed survey found. Adoption climbed to 61 percent of service firms, from 40 percent a year earlier, and reached 51 percent of manufacturers; about 15 percent of service firms hired fewer people and about 13 percent hired more. Just over a third retrained staff, yet median AI use covered only 17 percent of workers in services and 7 percent in manufacturing.
Banca d'Italia — A Bank of Italy occasional paper by Tiziano Ropele and Alex Tagliabracci finds that AI adoption lifts firms' labour productivity and profitability. Using the bank's survey of industrial and private non-financial service firms with 50-plus staff, matched to balance sheets and worker data, it sees a shift to higher-skilled jobs with total employment holding; just over 10% used AI in 2024.
In a randomised field experiment across 66 firms and 7,137 knowledge workers given a generative AI assistant inside their existing email, meeting and writing tools, the 80% of treated workers who used it cut email time by two hours a week over the final three months of the six-month trial and worked less out of hours. Researchers found no shift in how much or what kind of work they did.
Ipsos (AI Monitor 2026) — Ipsos CIESMORI's Bolivia results in the Ipsos AI Monitor 2026 show 70% of workers reporting time saved through AI at work over the past year, eight points above the 62% global mean. Some 69% of Bolivians report understanding AI, 57% feel excited about AI products and 51% uneasy about workplace AI.
Microsoft (Global AI Diffusion Report Q2 2026) — Costa Rica leads Latin America and the Caribbean in generative AI adoption: 29.7% of 15-to-64-year-olds, per Microsoft's Global AI Diffusion Report for Q2 2026, up from 28.5%. The nation ranks 32nd globally, beating the 18.8% world figure plus Colombia (25.9%), Panama (24.2%), Mexico (21.4%) and Brazil (20.2%) on AI usage.
Statistics Norway (SSB) — Statistics Norway (SSB) counts nearly five in ten enterprises with at least ten employees using AI technologies in 2026, against three in ten in 2025, a leap in business AI adoption. Firms of 100 or more staff reached eight in ten, up from six; those with 10-19 staff reached four in ten, up from three. Enterprise AI use peaks at nine in ten in information and communication.
INSTAT (Institute of Statistics of Albania) — INSTAT puts AI adoption in business at 12.4% of Albanian enterprises with 10 or more employees during 2026, against 9% in 2025. Corporate AI use at firms above 250 staff climbed from 24.8% to 28.1%. The 2,384-enterprise sample is the first under the revised NVE Rev. 2.1 classification.
Deloitte UK — Deloitte polled 25,000 people on workplace AI adoption and found 63% of workers already use such tools on the job. Nearly one in ten UK employees have tried software the employer forbids or would frown on. A similar share fear managers believe AI at work could replace them.
SAP and Oxford Economics — 66% of Canadian organisations are piloting agentic AI, yet 97% admit they are not fully ready to deploy and govern it, SAP's Value of AI Report by Oxford Economics finds (2,600 business leaders in 13 countries, 200 in Canada). Among Canadian leaders, 70% either agreed or could not say whether they deploy AI agents faster than governance keeps pace, and 73% struggle with incomplete or inconsistent data.
Competitiveness Enhancement Fund for Target Industries, Thailand — Thailand's Competitiveness Enhancement Fund for Target Industries cleared 55 projects for AI in manufacturing, automation and digital technology, including 48 business transformation schemes worth 3.54 billion baht in investment against 1.663 billion baht of state backing. Seven workforce projects will train 10,180 workers. Ekniti Nitithanprasert chaired the subcommittee, linking industrial upgrades to digital transformation.
Xebia — One in four Benelux organisations run AI agents in production, per Xebia's Data & AI Monitor 2026/2027 Benelux Edition (56 questions, 415 professionals, covering readiness, governance, sovereignty and trust) — part of a broader AI agents in the workplace shift across the Netherlands and Belgium. Assistant adoption climbs much higher, reaching 74%, yet faith in the technology hasn't caught up: merely 21% rank machine output alongside a human colleague's, while only 24% believe they can fully monitor what AI actually produces. Dependence on external vendors runs deep as well, with 81% describing themselves as reliant and 69% anxious about unsanctioned tool use spreading among staff.
Tieto Austria and TQS Research & Consulting — Employers stay in the dark: research from Tieto Austria and TQS Research & Consulting puts covert workplace AI adoption at 17% of Austria's workforce, drawn from 795 employed people among 1,000 polled. Company-sanctioned tools alone satisfy only 23% of staff, whereas distrust of outputs and absent policy push 42% to skip AI use at work entirely.
Ministry of Economy and Innovation (Albania) — Economy and Innovation Minister Delina Ibrahimaj told Tirana's Second Innovation Forum that AI adoption covers just 1.8% of Albanian enterprises, or 9% counting firms able to use it as well as those already doing so. Productivity across Albanian businesses sits at roughly a third of the EU average, and small and medium firms — 99.8% of enterprises, about 80% of jobs and 75% of value added — lag furthest behind. Closing that gap, Ibrahimaj argued, is exactly what the technology could help do.
Bpifrance and the Direction générale des Entreprises — Over 35,000 companies in France engaged Bpifrance's Osez l'IA programme in year one, aided by 615 AI Ambassadors across 20 regions and 13 sectors. AI adoption in French business trails Paris's 2030 goals: just 70 Diagnostics Data IA assessments have begun against a 1,000 target, toward full uptake at large firms, 80% at SMEs and mid-caps, and 50% at micro-enterprises — AI uptake among SMEs still short.
Information and Decision Support Center, Egyptian Cabinet — The Egyptian Cabinet's Information and Decision Support Center found 65 per cent of respondents knew of AI chatbots such as ChatGPT and Gemini in 2026, up from 22 per cent in 2023. Among those aware of AI, 26.3 per cent use ChatGPT and 10.9 per cent Gemini, chiefly for work (51.7 per cent), personal matters (38.3 per cent) and study (32.3 per cent). Generative-AI habits were surveyed from 15 to 28 July.
Colmena66 (Puerto Rico Science, Technology and Research Trust) — Colmena66, the Puerto Rico Science, Technology and Research Trust's entrepreneurship programme, finds AI adoption among entrepreneurs at 53.7%, with 19.3% more planning to follow, in its State of the Business Community 2025 report, presented at the Boricua Emprende Fest on 4 September. Business AI use centres on innovation, process automation, and marketing and data analysis. It rests on a convenience sample of 555 entrepreneurs gathered from September 2025 to May 2026, so it does not represent all Puerto Rican entrepreneurs.
World Bank — The World Bank puts AI adoption in Thailand at about 12% of Thai businesses, launching its Building Thailand's Future Today report at the Bangkok Business Summit. Vice-president for East Asia and Pacific Carlos Felipe Jaramillo noted the one-in-eight rate despite an established electronics manufacturing base and fresh data-centre investment. The report ties the 2037 high-income target to raising it, particularly for small and medium-sized firms.
Statistical Service of Cyprus — AI adoption in Cyprus hit 9.3 per cent of firms in 2025, from 2.6 per cent in 2021, the Statistical Service found. Size drives the gap: 7.7 per cent among small firms with 10 to 49 staff, 15.3 among mid-sized ones, 35.1 at employers of 250-plus. Business use of AI was measured in selected sectors among firms of 10-plus staff.
Oficina Nacional de Estadística (ONE) — The National Statistics Office (ONE) puts business AI adoption at 35.0% of internet-connected formal firms in the National Survey of Economic Activity (ENAE) 2025. Corporate AI use peaks at 52.0% among employers of 250-plus staff, against 46.1% at firms of 50 to 99.
NECTEC — AI adoption in Thailand has hit 53.7%, a NECTEC poll of 574 Thai bodies shows, though maturity across five dimensions - governance, technology, people, data infrastructure, strategy - averages only 'Aware'. Thailand's national electronics laboratory, releasing AI Readiness 2026, urged faster progress on governance and technology for effective artificial intelligence deployment. A dashboard follows 29 September 2026.
Federal Reserve Bank of St. Louis — The share of US workers using AI for their jobs rose from 33% to 45% between August 2024 and May 2026, and adult AI use from 45% to 62%, St. Louis Fed economists Bick, Blandin, Deming and Schumacher found. Yet AI use at work runs shallow: workplace adoption passes 50% in only 40% of occupations and fewer than 3% of tasks.
Endeavor Data Unit / Xertica.ai — Endeavor's data unit, backed by Xertica.ai, finds in its Colombia AI Pulse study that 61% of Colombian companies call AI a strategic priority, yet only 37.7% have made it a significant part of daily operations. On AI adoption in Colombia, 82% already see measurable impact and 72% raised AI budgets last year; the country scores 45.6 of 100 on the AI Sophistication Index, near Mexico's 45.1.
SONDA — SONDA's Tech Trends study puts some degree of AI adoption among Chilean companies at 88%, up from 16% in its previous edition. Of those, 40% are exploring or piloting, 30% apply it to some processes, 13% treat it as a strategic pillar and 5% have woven it through the whole operation. SONDA Chile general manager Javier Larenas warns of pilots that never scale; 53% have only general AI guidelines, and the sample size is unpublished.
Andorra Digital — Andorra Digital's 2026 Digital Divide Study puts AI adoption among Andorran companies at 25%, up on 17% in 2022, third in the European Union. Smaller firms trail at 24% despite forming 80 to 90% of the base, 12% carry ICT staff, and 96% of people go online daily. AI uptake here rests on 2024 replies from 800 residents and 678 companies.
Business Association of Georgia — BAG's survey of 600-plus Georgian firms puts AI adoption in Georgian business at 84%, spanning firms researching, piloting or running the technology. 13% report full integration, 24% are piloting, and 47% are still planning. Marketing and content generation led use cases at 20%, with finance and customer support each at 15%; fieldwork ran 9-25 June 2026, reflecting artificial intelligence use among Georgian companies.
ČSOB — Some 74 percent of firms surveyed for ČSOB's Index of business expectations report AI adoption in Czech business, yet thorough use reaches only 17.8 percent, with 38 percent dabbling, 19 percent confining it to back-office paperwork and 15 percent calling the technology a danger. Sectoral spread is wide: consulting and creative services stand near 57 percent and professional and IT services at 52.5 percent, well above accommodation on 27 percent and agriculture, forestry and fishing on 11 percent. Asked about the coming six to twelve months, 70.5 percent of consulting and creative houses, 57.4 percent in HR services and roughly 43 percent in construction intend fresh artificial intelligence projects, across 400 firms and entrepreneurs polled.
Awitec — Awitec's Baromètre IA Martinique 2026 polled 130 company decision-makers on AI adoption in Martinique businesses, a snapshot of generative AI use among company decision-makers. 50% had only tried generative AI once or twice out of curiosity, 33% acted as strategic thinking partners with the tools, and 17% reached expert level in automation and agentic AI. 89% still lack a working grasp of EU AI Act requirements.
Universite Technologique du Congo — 125 semi-structured interviews in Kinshasa, Lubumbashi and Goma reveal youthful Congolese running micro-businesses on AI in the informal economy: ChatGPT and other large language models, picture generators and chatbot apps shoulder translation, logo work, automatic customer replies, marketing copy, stock ledgers and crop planning. Delphin B. Kyubwa of the Universite Technologique du Congo credits swapped tips in WhatsApp groups, YouTube clips and tinkering, never a classroom; his coinage is compressed professionalization. Creaking infrastructure throttles informal-sector AI use worst: pricey bandwidth, erratic current.
Govern d'Andorra — Andorra's Business Digitalisation Programme, open from 8 April, exhausted a €500,000 pot on 178 firms and 368 digital solutions in 2026. AI adoption in Andorran business topped demand with 85 requests. Micro firms drew 69% of grants, small ones 27%, mid-sized 3% and large 1% of this AI uptake.
KPMG Baltics — Caution rises with size: about half of firms turning over 20 million euros or more have stood up local artificial intelligence, double the norm, which KPMG lays to security and data-protection nerves — a distinctive shape to AI adoption in Estonian enterprises. Across respondents, 69 percent confine AI to a limited slice of staff, though one large firm in five is past 50 percent. Gains run high: 93 percent cite faster throughput, 84 percent better independent problem-solving. Norstat polled firms above five million euros of turnover whose people use such tools, for KPMG Baltics OÜ, whose senior IT consultant Martin Käärik weighed the enterprise AI uptake.
Statistics Iceland — Iceland's national statistics office finds nearly half of firms already deploy some kind of automated software, according to a first-of-its-kind poll into business technology uptake. Statistics Iceland's inaugural enterprise survey, run in 2025 and released March 2026, puts AI in business adoption at 48%, with 50% using specified AI technologies. Just 25% apply it to defined tasks and only 14% have a formal strategy, though text-analysis leads use at 45%. Adoption among enterprises reaches 81% among firms with 250-plus staff.
rabota.by research centre — Neural networks already run inside business processes at 21% of firms polled by the rabota.by research centre, with 13% intending to adopt artificial intelligence. The centre questioned 2,199 service users in November 2025, from specialists to owners, mapping AI in Belarusian business. Positive feeling towards workplace neural networks reaches 56%, and 43% already draft text, answer professional queries or produce content. Employers apply them mainly in marketing analytics, planning, forecasting, production, staff training and customer service.
ilik konsulentfirma & Grønlands Erhverv — A poll of local firms by ilik konsulentfirma and the Confederation of Greenland Industry (Grønlands Erhverv) puts business AI adoption in Greenland at 85%. Staff lean chiefly on ChatGPT, Claude, Mistral and Copilot for drafting copy, running analysis, and shaping marketing or social posts. Nearly all, 91%, expect their AI use in Greenlandic companies to grow further, yet thin knowledge and weak skills still stand as the top obstacle.
Nearly nine in ten organisations report using AI in at least one business function, up from 88%, and about 72% use generative AI in at least one, McKinsey's State of AI: Global Survey 2026 finds. Adoption is near-universal, though the respondent panel is self-selected.
21.8% of U.S. businesses reported using AI in some business function in the two weeks to 26 July 2026, and 25.9% expected to within six months (Census Bureau BTOS, 13 August 2026). Use climbs with firm size: about 20–22% under 50 employees, 30.3% at 100–249 and 41.5% at 250-plus.
A nationally representative U.S. Census study of 117,000-plus firms (Nov 2025–Jan 2026) found 18% using AI in a business function, 32% employment-weighted. AI adoption is shallow: 57% of adopters use it in three or fewer functions, chiefly sales and marketing (52%), strategy and business development (45%) and IT (41%); 66% of users only augment existing tasks, and 2% of firms cut AI-related jobs.
UK AI adoption has spread without deepening. Among businesses with 10-plus staff, self-reported use of at least one AI technology rose from around 12% in late 2023 to around 35% by June 2026, and 49% at firms of 250-plus. Yet adopters' average number of AI technologies grew only from about 1.4 to 1.6, and just 10% report extensive AI use.
Canadian business AI adoption reached 19.2% in Q2 2026, triple the 6.1% of Q2 2024, counting use to produce goods or deliver services over the prior 12 months. Rates ran from 42.3% in information and cultural industries to 4.5% in agriculture; 44.4% of firms changed training or staffing because of AI use. Cybersecurity or privacy (13.4%) and cost (10.6%) led the barriers.
Among nearly 6,000 senior executives at US, UK, German and Australian firms, 69% say their firms actively use AI, yet nine in ten say it has not affected their firm's employment or productivity in the past three years. For the next three years they expect modest effects: productivity up an average 1.4%, output up 0.8% and employment down 0.7%.
About 55% of U.S. workers say they have used AI at work for at least one of 11 tasks, per a Census Bureau household survey fielded March 2026. Among last week's users, 31% saved one to two hours, 25% under an hour, 15% three to four hours and 15% over four hours, while 10% saved no time and 3% needed extra time.
Artificial Intelligence Foundation — The Artificial Intelligence Foundation's 2026 survey of 228 companies in more than ten sectors lifted its AI adoption in Taiwan score to 43.62 from 36.77 a year earlier. Firms at the Ready AI or Scaling AI stages made up 47.8 per cent, and those still classed Unknowing AI fell to 26.8 per cent. ICT leads while mining, fishing, forestry, farming, construction and engineering trail well behind.
Microsoft AI Economy Institute — AI adoption in Lebanon hit 27.3% of the working-age population during the first quarter of 2026, on the Microsoft AI Economy Institute ranking — above 19.1% worldwide and 22% across Arab countries. Lebanon sits 33rd of 147 nations, fifth of 16 Arab states. The gauge tracks generative AI use, derived from anonymised telemetry Microsoft harvests within its ecosystems and calibrates for internet penetration and device share.
Microsoft AI Economy Institute — AI adoption in Bosnia and Herzegovina reached 22.1% of the working-age population in the first quarter of 2026, 2.6 points above 19.5% for the second half of 2025 and 18.2% for the half before. Worldwide the share moved from 16.3% to 17.8% over those months, on the Microsoft AI Economy Institute count. Its gauge tracks generative AI use among 15- to 64-year-olds, from pooled, anonymised Microsoft telemetry corrected for operating system, device-market share, internet reach and population.
Narrative Research Bermuda — Narrative Research Bermuda puts weekly AI use at work at nearly six in ten employed residents, and half of residents use AI weekly privately. 78 per cent gave strongest agreement to regulating AI; 62 per cent fear for jobs, 58 per cent flag AI-related dangers and 57 per cent dread leaks of confidential or sensitive information, while 33 per cent voice excitement. Fieldwork ran 5 December 2025 to 4 January 2026 across 248 residents aged 18 upwards on its Bermuda Voice panel, where $75,000-plus earners showed heavier workplace AI adoption.
Microsoft AI Economy Institute — AI adoption in the Central African Republic reached 8.7% of the working-age population in the first quarter of 2026, 0.9 points above 7.8% for the second half of 2025 and 7.0% for the half before. Worldwide the share moved from 16.3% to 17.8% over those months, on the Microsoft AI Economy Institute count. Its gauge tracks generative AI use among 15- to 64-year-olds, from pooled, anonymised Microsoft telemetry corrected for operating system, device-market share, internet reach and population.
Microsoft AI Economy Institute — AI adoption in Chad stands at 8.7% of the working-age population for the opening quarter of 2026, a gain of 0.9 points on 7.8% across late 2025 and 7.0% earlier that year. The world average climbed 16.3% to 17.8%, the Global South 14.1% to 15.4%. Microsoft's AI Economy Institute derives generative AI use among 15- to 64-year-olds from pooled, anonymised telemetry, corrected for handset and operating-system mix, connectivity and headcount.
K&H Innovation Index — Corporate AI adoption in Hungary has reached a high in the K&H Innovation Index: 58 percent of firms turning over at least 300 million forints run some AI-based solution, and those building their own rose from 24 percent a half-year earlier to 36 percent. Phone interviews with 360 innovation managers ran from 14 April to 6 May. Thin skills, plus data protection and confidentiality worries, still brake adoption, cited by 66 and 52 percent of would-be AI users.
Eurostat / Statistics Portugal — Eurostat found AI in business use climbing to 11.5% of Portuguese firms with 10 or more staff in 2025, a 2.9-point rise since 2024. Written language analysis topped the applications list at 59.4%, ahead of image and audio generation at 50.9% and text or code generation at 45.6%. Uptake nonetheless lags the EU's average of 20%, showing steady but below-average AI adoption among firms in Portugal.
Eurostat — Bulgarian enterprise AI adoption hit 8.55% in 2025 among firms with 10+ staff, Eurostat reported, up from 6.47% in 2024 and 3.62% in 2023. It still trails the EU's 19.95% average, far behind leader Denmark's 42%, showing weak artificial intelligence use among businesses.
Taquion Group / RESTART / iplan / GCBA — Taquion Group's Monitor Nacional de Inteligencia Artificial 2025 surveyed 1,301 adults and found personal AI adoption in Argentina outpacing workplace AI use: six in ten use AI tools daily, versus 43% whose employer has adopted it. Rollout stays rare at 6%; only 31% of firms run a digital transformation unit, and 5% require training.
European Investment Bank — European Investment Bank's 2025 survey of 482 Czech firms puts generative AI adoption at 48%, above the 37% EU average and leading Central and Eastern Europe. Adopters' gen AI use: internal processes (67%), marketing and sales (45%), customer service (41%).
Statistics Finland — Finnish enterprise AI adoption jumped 14 percentage points year-on-year to hit 38% in spring 2025, Statistics Finland found. Information and communication businesses topped the ranking at 80% penetration, while larger organisations — headcount above 100 — reported 68% were using AI technologies.
Hong Kong Productivity Council (HKPC) — Hong Kong Productivity Council's AI Readiness in Workplace Survey 2025, drawn from around 800 local companies, found that workplace AI adoption already reaches 88% of staff for daily tasks such as customer service, data analysis and marketing. A further 92% of organisations intend to build AI into formal workflows, with 24% targeting full rollout within a year, yet only 45% currently operate official AI platforms. Talent shortages and data privacy concerns remain the biggest barriers, and 54% still lack a complete AI governance framework, showing AI adoption in the workplace has outpaced structured oversight.
Datacom — AI adoption in business has surged across New Zealand, with Datacom's 2025 State of AI Index putting the share of organisations using some form of AI at 87%, up from 66% in 2024 and 48% in 2023. Among the 200 senior business leaders polled, 88% of those already onboard flagged a measurable positive impact, and a fifth, 20%, logged productivity leaps of at least 25%. Full-scale rollout still lags behind, though: a mere 12% have gone enterprise-wide, and nearly half, 46%, remain stuck testing the waters on this pattern of AI usage among organisations.
SHRM-PR and Aon — Society for Human Resource Management Puerto Rico chapter (SHRM-PR) and Aon's joint survey shows 54% of Puerto Rican businesses now run pilot projects testing AI adoption in Puerto Rico, yet just 4% have built structured programmes managing the technology's rollout. This gap between quick experimentation and formal AI pilot projects points to a business sector still working out how to scale early trials into lasting systems.
Global corporate AI investment hit US$581.7 billion in 2025, up 130% year on year, including US$344.7 billion of private AI investment (up 127.5% on 2024), with generative AI taking nearly half. This replaces the ledger's 2024 figure of US$252.3 billion (private investment up 44.5% YoY) from the 2025 AI Index.
AI adoption in EU enterprises with 10 or more employees reached 20.0% in 2025, 6.5 percentage points above 2024's 13.5%. Uptake climbs with company size (17% small, 30.4% medium, 55.0% large) and spans 42.0% in Denmark to 5.2% in Romania.
NHO (Confederation of Norwegian Enterprise) — AI adoption Norwegian businesses hit 55 percent, more than double the 24 percent of 2023, per Samfunnsøkonomisk Analyse's survey for NHO. One in five firms rank as front-runners, embedding artificial intelligence use in Norway's corporate sector into core operations for outsized productivity gains.
Linking AI-adoption survey answers to administrative business microdata, Statistics Canada found the raw productivity premium of AI adopters mostly vanishes once other firm traits are controlled for. Adopters' labour productivity was 16.8% above non-adopters' at baseline, 10.2% after controlling for prior productivity, and 5.1%, statistically insignificant, after adding complementary capabilities: robotics, cloud computing, data analytics, R&D and ICT training. Built on 2019 and 2021 survey waves, it predates generative AI.
Regulation (EU) 2026/1744, the 'AI Omnibus', took effect on 27 July 2026, delaying AI Act high-risk duties: stand-alone Annex III systems from 2 December 2027 and AI in Annex I regulated products from 2 August 2028, instead of 2 August 2026. Its AI Act simplification gives small mid-caps relief once limited to SMEs, widens regulatory sandbox access with an EU-level sandbox and eases database registration for exempted systems.
The Adecco Group — The Adecco Group is giving 27,000 employees across over 40 countries Salesforce's Agentforce Coworker, bringing agentic AI in recruitment into routine work after UK and France pilots. Announced from Zurich, the tool assists selling, hiring and dealings with clients and candidates. Recruiter-facing AI agents already run in ten markets producing 50% of group revenue.
Universidad Nacional de La Plata (PROLAB) — Universidad Nacional de La Plata has opened Nodos de Empleo, free AI job matching run by its PROLAB careers service. Advanced algorithms weigh each professional profile against vacancies employers post, linking jobseekers, graduates, firms and community bodies without an intermediary. Talentia furnished the technology; PROLAB lent reach and local standing. Any organisation may raise a node, paying nothing to build one.
Banco Hipotecario del Uruguay (BHU) — Banco Hipotecario del Uruguay (BHU) is screening 2026 administrative internship candidates through AI job interviews on the MeetPIA platform. The eliminatory video interviews run about 45 minutes without pauses between 11 and 15 September, scoring up to 30 points with 15 needed to pass. MeetPIA chief executive Martín Liguori says the system assesses and suggests while a person always decides.
Mitsubishi Electric — BAKUTAN OS now underpins HR records for roughly 15万 (150,000) staff across 200-plus Mitsubishi Electric group firms, built by University of Tokyo spin-off BAKUTAN for this AI HR data integration push. Agents translate disparate HR formats into one shared standard, cutting integration effort by 60% or more. A pilot also tests algorithmic matching of rising leaders to roles, part of a shift toward AI-powered talent management.
Shoprite Group — Shoprite Group deployed generative AI in retail for roughly 12,000 workers via Shoprite AI, a homegrown tool with a ChatGPT-style interface that automatically routes each request toward whatever model works best and cheapest. CTO Chris Shortt framed it as an operational rollout, not a pilot, touching general workplace output and merchandising judgement calls including category management, pricing and replenishment.
Maya — Fintech Maya has put AI in HR operations alongside employee self-service, so staff produce their own certificates of employment and similar documents while managers handle promotions and transfers themselves. AI checks the company's BIR Form 2316 filings for inconsistencies and supports recruitment through market analysis and candidate sourcing. People group head Charee Lanza says technology clears away tasks employees should manage without waiting for HR.
NTT East Group — Roughly 30,000 NTT East Group staff now work under Mikiwame AI, an AI-based employee monitoring software covering personality checks, wellbeing surveys and management tools group-wide. Leadingmark built the engine; NTT HumanEX advises on turning its output into development moves, an AI employee wellbeing platform in practice.
Toyota Technical Development — AI resume screening has reached Toyota Technical Development: JAPAN AI HR filters roughly 800 applications yearly, once a single staff member's job. Weighted scoring replaces ad hoc judgement, giving the Toyota Group firm a uniform yardstick, while managers get the system's verdicts for the screening that follows.
Mitsukoshi Isetan Human Solutions — A 40% drop in interview workload followed Mitsukoshi Isetan Human Solutions' adoption of JAPAN AI HR for AI interview transcription across roughly 180 to 200 annual hires. Fixed-criteria scoring and automatic minutes replaced its dual-interviewer panel with one reviewer plus the system, standardising candidate evaluation company-wide.
In Mobley v Workday, a US federal court let a nationwide ADEA age-discrimination collective action proceed over Workday's AI hiring-screening tools, for applicants 40 and over rejected since September 2020. The flagship US test of AI hiring bias, it asks whether a vendor can be liable as an employer's 'agent'; opt-ins closed in March 2026 and core discrimination claims advanced in June 2026.
LinkedIn has launched Hiring Assistant, billed as its first AI 'agent', which sources candidates and builds shortlists autonomously, freeing recruiters for relationships. Early enterprise user Certis, a security firm, reported recruiter productivity up 60 to 70 per cent and shortlists produced in minutes instead of days, per its chief human resources officer.
Ministry of Labour — Qatar's Ministry of Labour now runs AI labour market forecasting through a revamped Labour Market Information System (LMIS), built jointly with the UN Economic and Social Commission for Western Asia. A central AI agent, drawing on domestic employment records and policy rules, generates sector-level job-demand projections across three strategic lenses. Staff quiz the agent directly for tailored real-time breakdowns split by demographic group.
Federal Authority for Government Human Resources (FAHR) — Federal Authority for Government Human Resources built Injazati, an AI performance management system rolling out UAE-wide from January 2026. A unified dashboard fuses OKRs, KPIs and core competencies, giving instant feedback and linking staff targets to national priorities. HR chiefs gain visibility through its AI-driven workforce analytics.
Department of Personnel Management — Papua New Guinea's public service now hires through an AI-powered e-recruitment system covering every government body, unveiled by the Department of Personnel Management at the Digital Transformation Summit 2025 in Port Moresby. Secretary Taies Sansan described an AI recruitment platform built on large language models that scans applicants' CVs, transcripts and credentials, aligns their skills against open roles, assembles candidate shortlists, and drafts offer letters plus contracts automatically. Trials start at the Department of Personnel Management and the Department of Information and Communications Technology, before an 18-month expansion brings every remaining national, provincial and health-sector office on board.
Cera — Cera, a leading UK home care provider, built Ami, an AI recruitment agent now live across its 10,000-strong workforce. Screening costs fell two-thirds and recruiters freed up two workdays weekly, since Ami conducts candidate interviews and screening alone. Cera also licenses this AI hiring tool industry-wide.
Bestjobs — Over 70,000 candidate interactions inside a single 60-day stretch: that's what Bestjobs, Romania's largest recruitment platform, attributes to its 24/7 AI recruiting agent Bestie. Working non-stop without human oversight, the system independently handled upwards of 10,000 interviews spanning video, text and audio formats, connecting with more than 17,000 distinct users while managing candidate screening and booking interview slots as an autonomous AI recruiter.
Growth Experts — Growth Experts fields staff queries, leave requests and system registrations through IBM watsonx Orchestrate, an AI agent in HR running chat support all 24 hours. Routine handling sped up 83%, manual input errors reached zero via integration, and foreign staff gained multi-language support. Accounting and general affairs are next for this watsonx Orchestrate expansion.
IPB University (Institut Pertanian Bogor) — AI talent management has arrived at IPB University, marking Indonesia's first public-campus rollout of a Talent Management System, built on ESQ Corp's TalentDNA platform and unveiled at the IPB Stakeholder Forum in ICE BSD, Tangerang. Hidden aptitude among students, lecturers and staff now surfaces through software rather than manual review, steering placement choices, scholarship awards and recruitment interest from industry partners — an early case of AI-driven talent identification in higher education.
LC Waikiki — LC Waikiki applies AI in human resources via six assistants: EDA (training announcements), KORAY (KPI guidance), SEDA (training evaluation), SERA (procurement simulation), Elsie ChatBot (24/7 payroll and leave queries), and MİYA, a recruitment-speeding avatar. About 500 staff joined an AI4FUTURE contest building RPA and agent tools — workforce automation piloted at the Turkish retailer.
Snappmarket — Snappmarket, the grocery delivery unit of Snapp group, has introduced an AI hiring assistant named Banafsheh to run first-round job interviews, built in partnership with startup Highlight. Trial participants called the process modern, transparent and motivating, citing lower anxiety and greater fairness than ordinary phone screenings. Snappmarket claims to be Iran's first firm hiring staff via this AI-driven interview screening, freeing HR teams from manual early-stage interview work.
ČEZ — ČEZ, Raiffeisenbank and Bageterie Boulevard now run first-round job interviews through an AI job interviews assistant from Czech-Slovak startup Talentiqa. Dozens of agents chat with applicants around the clock, narrowing candidates to a shortlist within a day. Talentiqa says the tool frees up sto hodin (one hundred) hours of recruiter time each month, a claim backing its fresh 25 million crown round from Purple Ventures and Venture to Future Fund for AI-driven candidate screening.
Livspace — CoHyre.AI cut Livspace's cost per hire by 30 to 35%. The Indian home interiors firm saw hiring cycles shrink 40%, cultural fit climb 20%, and roles once stuck at 45 days close in 15 to 20 days — an agentic AI recruitment platform delivering an AI hiring platform upgrade.
XpressJobs — XpressJobs unveiled Sri Lanka's first AI-powered recruitment assistant, texting pre-matched candidates via WhatsApp, fielding thousands at once. Interaction hit 85% in John Keells Retail Sector's launch, an early sign of AI-powered hiring tools taking hold locally.
Esperia Group — Wages for over 1,250 workers at Esperia Group, based in Rhodes, now flow through an AI-powered HR platform running on Entersoft-SOFTONE's WorkLife HCM system. Tips, overtime, benefits and clock-in records are managed through automated payroll processing, while attendance figures pass automatically to ERGANI II under Greece's Digital Work Card scheme.
Ansa McAL Ltd — Trinidad and Tobago's Ansa McAL claims a first for the English-speaking Caribbean: the earliest large-scale enterprise running AI in human resources throughout its business, joining a generative AI assistant to its human capital management setup via SAP and HRIZONS. Workers can pose HR questions conversationally and get tailored study and career pointers, backing a plan to renew infrastructure and boost how staff experience an AI-driven HR platform.
City of Sydney — A skills framework for 2,000 City of Sydney staff took six weeks with SAP Business AI, an AI in HR skills management case, versus months before. Discovery Consulting merged hiring, growth, and appraisals into one system, cutting paperwork 25%. The AI-driven skills taxonomy updates itself as needs shift.
COSAPI — Peruvian builder COSAPI now screens, trains and evaluates job candidates through COSAPIA, an HR platform marking one of the earliest large-scale uses of AI in human resources recruitment nationwide. A WhatsApp chatbot answers queries, cross-checks paperwork against official databases, and fires off notifications and reports unattended — automated candidate screening replacing manual work.
Biedronka (Jeronimo Martins Polska) — Biedronka, Poland's biggest supermarket group, has rolled out jAIn, an AI leadership development tool for HR building tailored growth plans for staff. A conversational assistant discusses each employee's goals, then drafts a plan on the 70/20/10 framework linked to EduAkcja, its e-learning system, for AI-driven personalised coaching.
Red Sea Global — Red Sea Global (RSG), a PIF-owned developer, is the Middle East and Africa North region's first firm running SAP's AI copilot Joule across human capital management, an early case of AI in human resources in real estate. Candidate screening, job description drafting and performance tracking now sit inside SAP SuccessFactors, covering roughly 10,500 staff. RSG will next extend the SuccessFactors AI assistant to ERP and procurement.
Adecco UK — Human interviews at Adecco UK & Ireland now happen only after an agentic assistant built on Salesforce's Agentforce platform has already messaged applicants by SMS or email, handling AI candidate screening automatically. Most of these automated recruitment chats, over half, land outside typical office hours. Applicants themselves gave the process an average score of 4.6 out of 5.
Bunnings — Bunnings is piloting AI in human resources to map career goals across 55,000 staff, pairing Workday Skills Cloud, built on the vendor's Illuminate AI, with HiredScore for AI workforce skills matching to internal vacancies. A chatbot separately handles queries on leave, policy and procedure so employees skip a human contact.
Anglicare Sydney — Hiring at Anglicare Sydney now takes six days, down from 42, after AI recruitment tools joined its Workday HCM system. Yvette McDonald credited the AI-assisted staff recruitment shift at the Workday Elevate summit, spanning 5,500 staff, onboarding and job coaching.
Ministry of Employment and Labor — Job seekers using Work-Net, the AI job matching platform run by South Korea's Ministry of Employment and Labor, secured over 87,000 placements across the first half of 2025 — an 84% jump against the equivalent stretch in 2024. Matching candidates against openings via analysis of resumes, work history and career goals, the system also produced a wage gap: workers placed through its recommendations took home 120,000 won more each month than those who landed jobs unassisted. Come September, the ministry adds four further AI capabilities to the platform, including an aptitude test built around the individual and a tool projecting employment probability, extending its AI recruitment matching capacity.
Cherkizovo Group — Polina AI, an AI recruitment assistant built by ITFB Group, launched at Cherkizovo Group in 2025, with six staff scanning CVs through an hh.ru browser add-on from day one. An LLM ranks files and flags leading candidates, trimming 10-15% off hours once spent on cold sourcing and paperwork. This AI hiring tool could soon spread across the whole business.
Infostud — Ten new AI recruitment tools now run on Infostud, Serbia's biggest job platform, built to speed hiring and cut screening workloads. Smart Sort ranks applications by relevance; other features summarise CV documents, draft job descriptions, generate interview questions and fill candidate profiles automatically, aiming for sharper employer-candidate matches through AI-driven hiring rather than guesswork.
Banco Galicia — Banco Galicia has embedded AI in HR processes across the organisation, from drafting job descriptions and setting performance goals to running interviews and internal learning programmes. According to the bank's head of HRBP and Talent, the tools help pinpoint skills gaps and guide employee development tailored to each role. This AI-driven talent management approach links recruitment, training and workforce planning under one system.
MediaMarkt Türkiye — Turkish electronics retailer MediaMarkt Türkiye cut hiring time by 78% using AI recruitment tools: AI'DA for automated pre-interviews and AI'M Hiring, an AI-integrated recruitment platform. In one year the AI-based hiring process screened 916 candidates, cut pre-interview time by 66%, hit 89% candidate satisfaction and filled 21 posts via AI'DA. Figures are the company's own, unaudited.
Manulife — Manulife has extended generative AI in insurance to 100% of its global workforce, and 75% regularly turn to ChatMFC, its proprietary GenAI assistant, alongside supporting training. Spanning Canada, the US and Asia, over 35 GenAI applications are already operating, with 70 further ones lined up for 2025. Nearly 200 machine learning engineers and data scientists underpin this generative AI adoption in insurance, pushing usage company-wide.
Al Zaid Al Quraishi & Brothers Company (AZAQ) — AZAQ, a Saudi holding firm spanning cars, consumer goods, logistics and property, now runs Elevatus's AI recruitment platform group-wide, folding AI in recruitment into daily hiring. A new branded portal automatically screens, sources and onboards job seekers using Elevatus's AI-powered hiring software. The rollout backs Vision 2030's diversification push.
SK C&C — Four hours: that is how long SK C&C's new AI recruitment platform now needs to sift a stack of job applications that used to take a week. Co-built with SK Telecom, A.Biz HR applies an AI capability test (AICT) gauging each applicant's technical grasp, problem-solving and readiness for the job, and now runs across entry-level and junior hiring at the South Korean firm's AI hiring system. Since an internal trial went well, bosses will next bolt on AI talent search and recommendation tools, due later this year.
National Oil Corporation of Libya — National Oil Corporation now filters CVs and job applications through Connect, an AI recruitment automation platform covering its Libya workforce. Employees receive individually matched training suggestions while the tool separately tracks performance outcomes. NOC's HR division co-developed this AI-driven HR platform alongside Med Oil Istanbul.
Chemist Warehouse — At Chemist Warehouse, staff can now get HR answers drafted automatically through AIHRA (AI HR Advisory), a shared-inbox agent handling roughly 300 weekly email queries sent by 30,000 employees and store owners across Australia — an early rollout of AI HR email management. Human advisors still check each draft before anything is sent out. Insurgence AI, a Microsoft partner, built the tool on Azure AI Foundry combined with Power Platform, and Chemist Warehouse reckons it now saves about 1,950 hours yearly, turning it into a blueprint other departments are copying for AI-assisted HR query handling.
JD.com — More than 100,000 Chinese job seekers have used JD.com's AI interview agent recruitment tool, built on its JoyAI large model, to screen for entry-level logistics and support roles. The e-commerce firm says the system now handles tens of thousands of first-round conversations daily, recognising over 20 regional dialects while flagging impersonation through digital-human hosts with layered verification checks — an AI-driven hiring screening approach. JD.com credits this with lifting hiring speed by 70% and scoring 91% accuracy, though these are the company's own unaudited figures.
Deutsche Telekom — Askt, Deutsche Telekom's internal generative AI assistant, now serves 50,000 active users handling roughly 500,000 monthly queries on HR, finance and procurement matters, part of a broader push into AI in human resources. Alongside it sit Growth Hub, offering tailored learning recommendations, and Nadja, a virtual coach rehearsing difficult workplace dialogues with managers. Even as these AI-driven HR tools spread, Telekom stressed that people, not algorithms, sign off on hiring and personnel outcomes.
Ministry of Labour (Oman) — Oman's Ministry of Labour has introduced Nadeem, billed as the nation's first AI career advisor, powered by generative AI. It guides job seekers through employment hurdles, matches them to tailored solutions, and feeds usage data into future job-matching and programme design, the AI-powered virtual career advisor's role within the ministry's Smart Management Project.
Apiux — Interview slots now sit in candidates' own hands as Apiux rolls out PoolRecruitIA's avatars for AI in IT recruitment across Chile. Automated note-taking, ATS updates and unbiased scoring trim evaluation time over 80%, pushing manual AI-driven candidate screening out of Apiux's pipeline.
Telenor Pakistan — Telenor Pakistan built an AI HR assistant chatbot in Microsoft Teams for quick staff answers on HR policies. Shown at its AI Day and AI Expo in Islamabad beside systems for service monitoring and smart energy management, it marks fresh use of AI in HR support.
Vietjet Air — Vietjet Air now runs AI recruitment for cabin crew hiring, becoming Vietnam's first carrier to fold artificial intelligence into staff selection. A pair of tools drives the workflow: AI Interview builds tailored questions from an applicant's background and rates language ability, teamwork and customer service, whereas AI Grooming Check inspects appearance. The Vietjet Aviation Academy hosted the debut round on 12 April 2025.
104 Job Bank (104人力銀行) — 104 Job Bank's AI System automates job-description drafting, company research and candidate matching, forming an AI recruitment automation tool for headhunters. Adoption reached over 70% of consultants within three months, shrinking case prep from about 10 hours to under 2 — an 80% gain that lifted division revenue by 6%, a result of the recruiting automation approach.
Chipotle has put Paradox's conversational AI hiring assistant, 'Ava Cado', in more than 3,500 restaurants across North America and Europe to handle candidate messages, interview scheduling and job offers. Announcing it in October 2024, the company said it expected in-restaurant time to hire to fall by up to 75 per cent.
Danmarks Nationalbank — Companies that take up artificial intelligence subsequently recruit fewer people than similar non-users, Danmarks Nationalbank finds in fresh work on AI and employment built from unique Danish records. The pullback falls hardest on younger and better-educated candidates; pay has not moved. Nothing yet points to a fall in national employment, and the central bank expects none soon. Director Signe Krogstrup argued that a flexible labour market and deep digitalisation leave Denmark well placed to gain from the AI jobs effect.
Stanford Digital Economy Lab — Employment of 22-to-25-year-olds in the most AI-exposed occupations is 19 per cent below the path of less-exposed peers, finds a revised version of the Stanford Digital Economy Lab paper Canaries in the Coal Mine. Using high-frequency ADP payroll data on millions of US workers through June 2026, Erik Brynjolfsson, Bharat Chandar and Ruyu Chen show the gap widening steadily since August 2025, driven by less hiring rather than more separations. Declines cluster where AI substitutes for people, while employment is flat or rising where it complements them; the authors call these early descriptive indicators, not causal evidence.
26% of job candidates trust AI to judge them fairly, and 52% think AI screens their application. Gartner surveyed about 3,000 candidates.
Recruiters using generative AI report saving about 20% of their workweek. It is roughly a workday.
PwC — PwC's 2026 Global Workforce Hopes and Fears Survey of 49,364 workers across 48 countries and regions puts AI use at work at 64% over 12 months, up 10 points. Daily generative AI users rose from 14% to 22%, while learning access fell from 59% to 51% in this May and June 2026 workplace AI adoption snapshot.
praca.by — AI job search runs deep among Belarusians, praca.by found: 29.5% do so routinely, 24.4% have tried neural networks on a vacancy at least once, and 20.7% mean to try. A quarter refuse outright. Of users, 63% build CVs and covering letters this way, 53% rehearse interviews, 47% dig into employers and 44% pick apart adverts. Only 12% would hand AI-assisted jobhunting the entire task.
LinkedIn gets about 11,000 applications a minute, up roughly 45% year on year. Generative-AI 'auto-apply' tools largely drive the surge, shifting the volume problem to recruiters.
HR governance lags adoption: 51% of organisations have no formal AI-use policy and 52% leave HR out of AI strategy (SHRM). Yet 62% expect AI to raise headcount in 2026, and only 7% expect a fall.
Deel — Roughly four-fifths of Chilean SMEs (80%) already embed AI in HR processes, a Deel and Criteria study of 300 firms shows. 92% plan broader use, led by attendance monitoring (29%), staff training (27%) and hiring (22%). 64% foresee raising budgets for AI-powered human resources management within 12 months.
Deel / Edelman DXI — Edelman DXI's research for Deel puts AI in HR adoption at 75 per cent among small and medium enterprises in Colombia. Payroll leads the use cases at 70 per cent, ahead of performance evaluation at 58 per cent, training and development at 57 per cent, and HR information systems at 52 per cent. Seven in ten SMEs plan to raise their HR automation spending.
Think Work — AI adoption in Brazilian human resources has reached 78% of companies, Think Work found, staff using such tools daily. A quarter of HR workflows (24% average) run with AI tools in HR departments support. The poll drew 203 respondents, firms of varying size, mostly Southeast region, 69% management roles.
SD Worx — SD Worx finds UK employers topping a 16-country European survey on AI in HR: 55% now fund such tools, a jump from 35% in 2024 and well past the 38% continental average. Britain's workforce likewise reports the heaviest routine AI use on the job, 31% versus 23% elsewhere. Training and development absorbs the biggest slice of this AI adoption in human resources (31%), with recruitment and HR support close behind at 28% apiece.
About 4 in 10 organisations (39%) now use AI in HR, with recruiting the top use case (27%). An earlier SHRM instrument put AI-in-HR use at 43% in 2025, from 26% in 2024; the 2026 figure uses a different survey base, so it shows the current level, not a fall.
37% of organisations are integrating or testing generative AI in hiring. A year earlier it was 27%.
State of Connecticut — Connecticut AI regulation advanced as parts of the Artificial Intelligence Responsibility and Transparency (CART) Act, Public Act No. 26-15, began: large frontier AI model whistleblower protection, AI subscription disclosure and AI definitions. Chatbot rules under this state AI law follow 1 January 2027. Employers using automated employment decision-making have until 1 October 2027 for a framework, must notify applicants and employees of AI, and cannot cite AI against discrimination claims.
State of California (Governor Gavin Newsom) — California AI regulation widened as Governor Gavin Newsom signed SB 947, curbing AI-driven dismissals and discipline, and SB 951, requiring disclosure of automation-linked layoffs. AB 1883 bans emotion prediction and brain-data collection, AB 1331 outlaws AI-powered bathroom surveillance and AB 1979 reserves patient-care calls for health professionals. Two other workplace AI rules, AB 2575 and SB 903, were vetoed.
Ministry of Labour and Solidarity (France) — Jean-Pierre Farandou's Ministry of Labour and Solidarity unveiled Impulsion IA 2027, pumping 80 million euros of new cash into 12 steps that shape France AI policy for workplaces. Extending Osez l'IA from summer 2025, the package brings Explore IA, a no-cost skills-check tool, CPF priority for AI courses and task forces coaching small and medium-sized firms. A labour-market AI observatory will monitor job effects, while 20 companies and sectors pilot fresh tech-focused social dialogue.
California Legislature — California lawmakers passed Assembly Bill 1883 on 30 August, sending Governor Gavin Newsom a measure barring employers from AI surveillance of a worker's individual emotional state or collection of their neural data. Violations would cost $500 each, with certain aircraft and product development for defence, military, space or national security exempt.
US Office of Personnel Management — The US Office of Personnel Management told agencies, in a memo from Director Scott Kupor, that AI in federal hiring is mostly not high-impact when used to draft job announcements, screen applicants, check files pre-offer or assess hiring metrics, as long as an official reviews it independently and adopts it. OPM plans AI features in USAJOBS, USA Staffing and USA Hire, but agencies need not wait for them.
The EU AI Act treats AI in recruitment and employment as 'high-risk', but the Digital Omnibus (Regulation (EU) 2026/1744, in force 27 July 2026) pushed those duties from August 2026 to 2 December 2027, and embedded product AI to 2 August 2028. Article 5 prohibitions have applied since February 2025: banned now, regulated later.
Octaplus Health — Octaplus Health has launched an AI customer engagement character, Xaida, designed to make health insurance easier to understand for younger, digitally connected members. Dr Dan Vincent Armooh said the teenage digital persona will feature across the firm's communications and digital content as a virtual interface for engaging members.
Taisho Pharmaceutical — A single day now replaces what once took two months: Taisho Pharmaceutical collapsed its entire consumer-persona workflow after bringing generative AI in pharmaceutical marketing to life through Google's Gemini, working alongside Dentsu. Collaborative strategy sessions driven by this AI-driven brand strategy approach reshaped how the Ri-Up hair growth brand reaches customers. Physical retail sales climbed 6% year-on-year afterwards, and EC revenue jumped 39%.
Bahari — Bahari posted Facebook images for the Coral World collection on 31 May built around AI-generated models, not photographed people. Shoppers pushed back, saying a label built on island culture should have booked real faces. General manager Carole Barnett replied that campaigns have leaned on local talent since 2014, and that AI fashion imagery suited this shoot because Coral World itself has closed.
Kontuur Leo Burnett — Every frame of "Must nimekiri", the campaign telling shoppers to check the Consumer Protection and Technical Regulatory Authority's register of traders who breached consumer rights, was an AI advertising production — television spot and digital placements alike. Urmas Villmann, the advertising entrepreneur behind it, told "Reedene intervjuu" the approach improves quality and cuts cost at once, one agency designer now specialising in prompting. Work once meaning film shoots billed in tens of thousands happens at a desk, though generative ad creation is no shortcut: coaxing a usable result from a prompt eats days, like a photo session.
M-PESA Ethiopia — M-PESA Ethiopia and Gebeya Inc. have launched the Dala AI Bundle, letting subscribers pay via M-PESA wallets for AI content creation in Ethiopia. Dala Studio covers app building, AI agent design, gaming, comics and material in Amharic, Oromo and other local languages, described as the first case of an African mobile-money operator folding such generative tools into a mainstream offering.
Google has put its Veo video model into Google Ads' Asset Studio and Demand Gen campaigns, letting advertisers make video ads of up to ten seconds from up to three images without leaving the ad console. Rolled out in March 2026 with Google's Nano Banana image tool, it brings AI-generated video creative into the ad-buying platform.
New York's Synthetic Performer Law, in force from 9 June 2026, makes advertisers and agencies conspicuously disclose knowingly using an AI-generated 'synthetic performer', a computer-made figure resembling no real person, in ads. Fines are $1,000 for a first violation and $5,000 after. Among the first US labelling rules for AI ad creative, it matches SAG-AFTRA's 2025 limits on synthetic performers.
A consumer backlash against AI-generated advertising has set in: Heineken, Polaroid and Cadbury market their work as 'human-made', and Apple added a 'Made by Humans' credit to its series Pluribus. H&M and Guess drew criticism for AI-created brand ambassadors. Authenticity worries now run so deep that major brands advertise not using AI.
Meta aims to make advertising fully end-to-end AI-generated by the end of 2026: an advertiser would give only a budget and a product image, and Meta's AI would create the imagery, video and copy, target across Facebook and Instagram, allocate spend and personalise each ad in real time. Chief executive Mark Zuckerberg set out the goal; the full system is not yet deployed.
Represent Communications — Represent Communications set up AI Studio, a Belgrade team for AI in corporate communications, digital campaigns and content production. It joins strategy, creative craft and automated pipelines around AI-assisted content creation, performance gains and workflow automation, under human watch, run by Lazar Jovanovic as Head of AI Studio.
JCDecaux Transport — SmartRetail's computer vision now tracks how riders move past ads in Hong Kong's transit network, part of a completed JCDecaux Transport trial at Tsim Sha Tsui Station. The system logged crowd flow, directional patterns, rider demographics and ad-screen attention, marking what JCDecaux calls Asia's first AI audience measurement effort in metro advertising. Ipsos verified the outcomes independently, underpinning this AI-driven ad analytics test.
Vodafone UK — Vodafone UK saves 20 weekly hours of manual banner work by adopting AI in marketing operations via Writer's enterprise platform. VOIS handles generative AI content production with the same tool, halving turnaround and lifting quality scores, per unaudited company figures.
Majid Al Futtaim / Carrefour UAE — Majid Al Futtaim lifted Sensodyne sales 68% by switching screens at 23 UAE Carrefour branches to AI retail media via Advertima's Audience AI, targeting viewers live. Over a 20-day A/B comparison, conventional loops managed 30%. Behind this AI-driven retail media sit its Precision Media Digital Screen Network, Grassfish digital signage and Adtrac orchestration.
KPN — A wintry scene starring mascot cat Louis marks KPN's debut fully AI-built Christmas advert, unveiled 4 December 2025 as an AI video advertising milestone. Google's Veo 3 tool generated each frame, with Dentsu Creative Amsterdam collaborating on the piece. The Dutch telecoms group termed it a controlled trial gauging what generative video tools can achieve for professional output.
Mechta.kz — Mechta.kz has launched Kazakhstan's first AI brand ambassador, Armanay, a virtual figure built in-house by its own designer. She will front adverts, video reviews and SMM for the electronics chain's digital transformation, part of a wider rise in AI virtual influencers across retail marketing.
Rohlik Group — One employee at Rohlik Group built the company's entire 2025 Christmas advertising campaign using generative AI, cycling through seven separate tools, over 1,500 prompts and roughly 3,000 generated assets across 80 working hours. This AI-generated holiday ad, an AI advertising campaign built without a studio, replaces a job that would normally need a full production crew and months of work.
Standard Bank — Seventeen genre-specific tracks (17) form the core of Standard Bank's "Undercover Ad" campaign, an exercise in generative AI in advertising that folds mobile app promotion into YouTube Music playlists instead of interrupting them. A single script fed the AI system that produced each track, tailored per genre so listeners barely notice the pitch. Digitally fluent, ad-averse audiences are the target, reached through AI-generated audio marketing rather than skippable spots.
Şok Marketler — A 90% cut in production time and cost came from Şok Marketler's new commercial, built entirely through generative AI tools from Google — Gemini 2.0, Imagen 4, Veo 3 and MusicFX — under a partnership with Google Türkiye in Turkey. Concept design, visual output, editing and the score were each machine-produced for the brand Amigo, marking the country's first AI-generated retail advertisement of its kind. The gain is an approach to AI-generated advertising the company itself reports, not an independently verified figure.
British Virgin Islands Tourist Board & Film Commission (BVITBFC) — British Virgin Islands Tourist Board & Film Commission now runs an AI trip planner on bvitourism.com via Mindtrip, an AI-powered travel platform. Type a question about restaurants, hotels, activities or ferries and get a tailored itinerary with maps and photos. Clive McCoy, Director of Tourism, says it spotlights small businesses and hidden gems islandwide.
Hyundai Italia — Hyundai Italia unveiled its debut AI advertising campaign for the Tucson Dark Line, broadcasting nationwide since 9 November 2025. Software transformed over 500 real photographs plus 30 minutes of raw video into an AI-generated TV spot carrying seamless Golden Hour glow and sweeping camera movement.
Super — 18 tools now sit inside an AI hub for advertising agencies built by Super, an independent regional advertising agency, covering in-house workflows, client work and third-party licensing. Influscan pulls together influencer-metric dashboards automatically, Visual Check flags grammar and design mistakes, and a Scoring Tool forecasts project results, backed by extras for scheduling, contract oversight and generative-AI prompting tips. Super thereby joins the wave of AI-powered marketing tools reshaping the sector.
Impulso (Grupo RD Saúde) — Impulso, Grupo RD Saúde's retail media unit in Brazil, cut campaign turnaround from three days to roughly five minutes with generative AI in retail media ad production at scale. The AI-generated ads roll out in stages: e-commerce creatives first, then email marketing and push notifications.
Felleskjøpet Agri — Felleskjøpet Agri's marketers debuted their first AI-generated advertising campaign, swapping generated animal imagery in marketing for genuine household pets. Agency Smeh and homemade AI software handled the entire build without outside help. Fresh internal rules now oblige clear labelling on any AI-generated element used.
Hong Kong Tourism Board — Hong Kong Tourism Board debuted its first AI-driven travel marketing push for the 2025 Hong Kong Wine & Dine Festival, built with dentsu Hong Kong and StackAdapt. It targets high-intent travellers via AI-generated insights, retargeting them on Agoda, Expedia Group, Skyscanner and Tripadvisor.
Dentsu — Dentsu APAC's new SaaS platform Zoyumi scores, produces and tracks creative output to deliver on-brand assets in under 30 minutes, with governance and brand-safety checks built in. Built in Singapore, the AI marketing campaign tool first ran for a global technology brand's rollout across several markets, targeting continuous, always-on execution over one-off bursts. It illustrates a broader shift toward automated ad creative generation with human oversight retained.
Visit Estonia — Estonia's official travel site runs an AI chatbot for Estonian tourism, built by Krabu Grupp, replying live on destinations, events and trips in five languages from any page of visitestonia.com. It serves as an AI travel assistant for millions of yearly travellers from over 100 nations.
Banco Popular Dominicano — Banco Popular Dominicano, with agency Pagés BBDO, has rolled out GIA, billed as the Dominican Republic's first AI-generated finance influencer for digital channels. Seven months of research, design and training on real sociodemographic data shaped the virtual influencer in banking, which posts educational financial content on Instagram as @gialosabe_.
SMG Swiss Marketplace Group — SMG Swiss Marketplace Group ran AI-generated ad campaigns for ImmoScout24 and AutoScout24, mixing synthetic moving images with real footage. Agency Freundliche Grüsse produced the AI advertising using Midjourney, ElevenLabs and Veo 3, generating over 500 assets for targeted distribution.
Chain Reactions Africa — Chain Reactions Africa, a Nigerian PR and marketing communications agency, has hired Ara by Prophet as a full-time team member in an AI robot workforce move, billed as the continent's first intelligent robotic insight system. It merges machine learning with cultural foresight to function as a live data analyst, cultural interpreter and strategy consultant on client accounts. The hire positions an artificial intelligence system as a named staff member rather than a background tool, reflecting a wider push toward AI-powered workforce automation in African marketing.
Mercado Libre and Samsung — GUT São Paulo applied generative AI in advertising for Samsung and Mercado Libre's Smart Week spot, machine-rendering image, video and music instead of filming. Jungle animals handle Samsung devices in the ad across Mexico, Brazil and Chile, the pair's third straight AI-made campaign.
Høie Retail — For AI marketing budget planning, Høie Retail relies on Odins.ai, a Norwegian-developed platform that also gauges campaign performance. Staff say the short trial left them feeling "naked without it". Behind it sits AI-driven Marketing Mix Modeling, rarely seen beyond big technology companies.
DAVIES MEYER — DAVIES MEYER produced EWE's AI-generated TV commercial for Connected TV entirely with generative software, combining Runway, Kling and Wondercraft so an animated cat could narrate the pitch for fibre-optic broadband. EWE, a German energy and telecoms firm, positions the ad as a repeatable model rather than a standalone project. A follow-up instalment built on the same AI-generated advertising campaign approach debuted on 14 July, confirming the method suits fast, recurring output.
Uludağ İçecek — Uludağ İçecek calls it Turkey's first AI-powered outdoor campaign: a generative AI figure dancing beside percussionist Burhan Öçal on billboards along Istanbul's Maslak-Levent corridor, promoting its Uludağ Şekersiz Limonata brand. The digital character moved to the drink's 9/8 rhythm jingle, blending live music with AI-generated advertising.
Carrefour — Carrefour's 2025 back-to-school push is its first AI-generated advertising campaign spanning France, Italy, Belgium and Romania. Publicis Conseil and Prodigious ran the process end to end on generative AI, plus AI-made print visuals. This marks growing AI-driven campaign production adoption.
Seguros Bolívar — Cinema audiences watching Jurassic World and F1 releases saw two adverts built with generative AI advertising for Seguros Bolívar's 'En Teoría' campaign. Rather than filming traditional scenes, the Colombian insurer generated virtual characters, backdrops and footage through Google Veo 3, an AI-generated advertising approach that trimmed over 100 million pesos from each spot's budget, by the company's own unverified count.
Super-Pharm — Super-Pharm, Israel's largest drugstore chain, unveiled an AI advertising production on 30 July, a commercial whose writing, imagery and soundtrack came entirely from machine tools, not human crews. Rona Lichten Ohaion named it the retailer's debut AI-generated television commercial.
Despegar — Despegar's AI marketing campaign "De Vacas" delivers six films built 100% with generative tools for image, video and music, made with independent shop Super. The set airs on Meta and YouTube in Mexico to drive hotel bookings, an early case of AI-generated advertising.
Taxfix — Tax platform Taxfix made fully AI-generated video ads for its Everything but Taxes campaign with Google's Veo 3. Each piece of generative AI advertising took two days from concept to finish at a fraction of conventional costs. Taxfix claims this was the first commercial Veo 3 deployment in Germany's advertising sector, and among the earliest by brands worldwide.
People's Leasing & Finance PLC — People's Leasing & Finance PLC now holds the title of Sri Lanka's first AI avatar for annual reports, courtesy of martech firm Enfection. Built on RAG and large language models, this corporate reporting AI fields questions on financial and sustainability metrics instead of static PDFs.
Kirin Holdings — Kirin Holdings put ChatGPT Enterprise AI deployment into its strategy, R&D and marketing units, covering patent searches, market studies, product creation and planning. Custom agents built for individual workflows come next on the OpenAI system, while a separate generative AI tool rollout called 'BuddyAI' already reaches roughly 15,000 staff inside Japan.
KFC South Africa — Nearly 10,000 people tried KFC South Africa's AI chatbot in fast food within its first week live. KAIIA (Kentucky AI Integrated Assistant), built by Cape Town studio Methodlab on Google's Gemini, chats in 11 South African languages about menu items and local culture — an example of conversational AI in fast food. Grant Macpherson supplied the count.
Hindustan Unilever — Small kirana retailers using Hindustan Unilever's Shikhar platform now build their own personalised promo clips through GenAI, and Samadhan gets stock to those shops by the next day, part of a wider push on AI in marketing retail and supply chain across the firm's India footprint. Sangam reallocates advertising budgets on the fly, working alongside separate tools churning out more than 6,000 branded assets each quarter. Skincare recommenders and virtual try-on features under Lakmé, Pond's and TRESemmé complete this AI-powered value chain deployment, having already drawn upwards of five million consumer trials.
Jung von Matt Spree — Jung von Matt Spree calls it Germany's first AI-produced TV commercial, made through generative AI advertising for Spreequell Kräuterbrause: a 30-second surreal piece titled Trink dir Berlin bergig built from AI visuals, AI voice-over and AI-composed music. The Berlin agency says this opened shots ordinary 3D work couldn't match within the same budget.
Coca-Cola Company — Campaign output up to 10 times faster, brand consistency intact: that's the pitch behind Project Fizzion, an AI design intelligence system The Coca-Cola Company built with Adobe in 2025. Instead of fixed rulebooks, it studies designer habits inside Adobe Creative Cloud and converts guidelines into adaptive, self-learning assets. Pilots already run the AI-powered brand design tool across 200-plus global brands.
Popeyes Turkey — Havas Istanbul crafted an AI-generated advertisement quick-service restaurant chain Popeyes Turkey debuted, using generative AI for every frame and sound. Unveiled 9 May 2025, it pushes two sauce-coated chicken flavours in fast-food marketing.
Bahrain Duty Free — An AI advertising campaign from Bahrain Duty Free marks its maiden fully robotic rollout, spanning airport signage, social feeds, radio and street hoardings. Bahraini slang meets English through AI-generated marketing, replacing agency writing. Bahrain Airport Company and Bahrain International Airport jointly built it.
Allee Shopping Centre (Budapest) — Budapest's Allee shopping centre fronted spring promotions with Alíz, a fully generative AI-generated marketing character shaped through human input. This virtual influencer campaign spread across the mall's screens, spotlighting goods, services and events. The rollout paired with a futurist exhibition in the atrium, spanning April to May 2025.
L'Oréal Groupe — L'Oréal Groupe cut AI marketing content creation time from weeks to days using Google's Imagen 3 and Gemini inside CREAITECH, its GenAI Beauty Content Lab. Prompts now yield brand-aligned visuals and eight-second Veo 2 clips across tens of thousands of products, applying generative AI content production in France.
HBL — HBL, Pakistan's largest bank, became the first bank in the country to deploy generative AI fan engagement, launching FanTunes to mark the tenth edition of the Pakistan Super League (HBLPSL). The tool lets supporters create their own personalised tracks using AI-driven music composition, tying the bank's brand to the tournament through an interactive, fan-generated experience. The initiative positions HBL as a pioneer in applying generative AI within Pakistan's banking sector for consumer engagement rather than back-office use.
EBM — English Biscuit Manufacturers (EBM) debuted Piper Ki Duniya under its Peek Freans Young Pipers' Club banner, billed as Pakistan's first AI-curated children's entertainment offering. Five Ramazan instalments wove narrative with automated selection tools, producing AI-curated kids' content for young viewers.
mediaReachOMD — MediaReachOMD's AI consumer intelligence platform, OMD Consumer Intelligence 2.0 (OCI 2.0), now covers 16 Nigerian markets, up from 7. Built-in generative AI handles analytical work, renders dynamic charts, and runs questionnaire simulation and rollout—pushing AI-driven consumer analytics past the tool's earlier reach.
Volvo — Volvo's comeback push in Saudi Arabia leant on AI in automotive marketing, unveiling the Kingdom's first fully AI-generated brand film, 'Come Back Stronger'. LION and Electromin, Petromin's e-mobility arm, left vehicles out of frame, anchoring the piece around one person's story instead. Machine-made production squeezed a months-long schedule into weeks.
Zong 4G — Zong 4G debuted Pakistan's first AI-generated television campaign, built for the ICC Champions Trophy. Machine-generated visuals and narrative replaced conventional production methods for the advert. The result signals wider adoption of AI-powered advertising content across Pakistan's telecom industry.
Pfister — Pfister, the Swiss furniture retailer, debuted 'Besser leben', a push built on AI morphing in advertising alongside creative agency Vasataschröder and production house Markenfilm Schweiz. Products reshape on screen through AI-driven visual shifts, tying sound sleep to a richer life. Digital feeds and social platforms now host the resulting clips.
Cadbury Pakistan — Fans of pop star Abdul Hannan could remake his track 'Iraadey' into a Valentine's Day message under an AI-powered personalised song campaign from Mondelez Pakistan's Cadbury brand. A gift-box QR code unlocked the generator. The tie-up brought AI-driven audio personalisation to Pakistan's confectionery marketing.
KDDI — KDDI's beta trials of an AI ad creative generation system, built with Recursive and Supership, cut digital marketing workload by 50% in Japan. A KPI prediction model trained on past campaign results flags likely winning banners, shaped to fit KDDI's brand guidelines and the 'au VISUAL IDENTITY' framework, as AI-generated banner ads.
VML South Africa — Half a year spent training on curated open-source pictures yielded Asili AI, VML South Africa's entry into the AI image generator for advertising space: rendering South African people and settings with genuine believability. Built atop Stable Diffusion's Flux model, the in-house system now equips the agency's clients and staff for commercial output, spanning digital ads to billboards. This route into AI image generation for advertising sidesteps stock photography that frequently distorts local context.
Österreich Werbung — A multilingual digital guide greets visitors for the Austrian National Tourist Office: its AI tourism avatar, shown at ITB Berlin, fields questions and pitches destinations across Austria. HeyGen and Mixed Reality AI drive it, standing in for staff on routine enquiries. Tourism boards can lease the same AI-driven travel assistant.
eMAG — AI video advertisements now come from eMAG, Romania's biggest online retailer, made with Veo, Kling, Runway and ElevenLabs. Written descriptions of characters, camera moves and scenes drive these generative systems, leaving eMAG's marketers in charge of concepts and direction. A spring 2025 fashion push, online-only, was the first result of this AI video advertising shift, after static graphics saw the same change in 2024.
Saint Lucia Tourism Authority — Saint Lucia Tourism Authority now runs AI tourism marketing to counter an 18% slide in UK arrivals after Virgin Atlantic and Tui axed their routes. Software sorts prospective travellers into segments such as wellness, heritage, adventure and honeymoon seekers, feeding tailored campaigns to each group. The push reflects an AI-driven visitor profiling effort aimed at rebuilding lost UK demand.
Swedish fintech Klarna says an internal Copy Assistant writes about 80 per cent of its marketing copy, alongside generative-AI image tools such as Midjourney, DALL-E and Adobe Firefly. In May 2024 it reported image-production costs down around $6 million and the image cycle cut to seven days from six weeks, crediting AI with roughly $10 million in yearly marketing savings and 25 per cent less external agency spend. Later cutbacks to its AI customer-service agents left marketing untouched, and these figures were never retracted.
Extreme Reach — 88% of agency and brand advertising professionals report using or piloting AI in creative production, nearly half daily, Extreme Reach's State of Ad Ops study of 400-plus US and UK professionals finds. Production and post-production dominate, led by visual effects and compositing (45%), then performance analysis and creative testing, and image creation, at 44% each.
Sber Bank Belarus — Sber Bank Belarus puts steady AI in marketing adoption at 79% across Belarusian firms, citing Minsk meetup figures. A mere 9% steer clear of artificial intelligence in marketing, blaming thin awareness over any missing use case. Others lean on content generation, digging up information, analytics, automating tasks and tailoring output.
US 'AI advertising' spending, ads around AI-powered results like Google's AI Overviews and chatbots, is forecast to more than double from US$32.03 billion in 2026 to US$68.25 billion in 2030 (about 2.1x). Over 80% goes beside AI content; in-chatbot ads stay under US$1 billion in 2026, reaching about US$5 billion by 2030.
89% of B2B marketers use AI to write marketing copy and 95% say their organisation uses AI-powered applications, in CMI/MarketingProfs' 2026 research (fieldwork mid-2025, n=1,015), against 81% team gen-AI use a year before. The wording moved from 'teams use generative AI' to 'org uses AI apps'; the 89% copy figure is the nearest like-for-like.
Nearly 90% of marketers report using generative AI at work. A September 2024 survey polled over 1,000.
Office of the Governor of California — Governor Gavin Newsom has signed Senate Bill 1050, Senator Angelique Ashby's AI advertising disclosure law: any video or audio ad using AI-generated performers to sell something must state that plainly, and non-compliant ads must be pulled. SAG-AFTRA, the bill's sponsor, hosted the signing of these synthetic performer labels at its headquarters.
Korea Fair Trade Commission — The Korea Fair Trade Commission's amended notice on substantiating labelling and advertising claims took effect on 3 September 2026, requiring prior substantiation for AI advertising claims and other new-technology claims about products. The extension for filing substantiation material shrinks from 30 days to 15, granted only on narrower grounds such as receivership, a merger, natural disaster or seized records. Firms failing to file can be ordered to pull the advertisement.
Kanpo Life Insurance — Kanpo Life Insurance rolled out bellSalesAI, a Salesforce data-entry agent from bellFace, to its 1,200-strong corporate sales workforce nationwide, using an AI assistant in corporate sales to pull structured details from calls. Post-meeting write-up time fell from roughly 50 minutes to about 10 — an 80% cut — with sharper record accuracy. Staff redirect the recovered hours to direct client engagement, an early case of AI assistant adoption in corporate insurance sales reshaping Japan's workforce routines.
The Property Franchise Group (TPFG) — TPFG is bringing AI in estate agency to the group's 1,900 franchised branches, boss Gareth Samples said, with 14 franchisees running roughly 25 offices on board, 50 more queued. Trial AI sales agents booked 25% more valuation appointments than usual methods, the company's own figure, across sales and lettings. Some 4,000 daily enquiries go unanswered in the UK network, the gap this property AI targets.
Honda Sales Operations Japan — Honda Sales Operations Japan now runs Agentforce, Salesforce's AI agent platform, inside its My Honda app to answer customers 24/7 with tailored replies. Applying AI in automotive sales, the tool lets drivers arrange test drives purely through chat, while an AI sales agent for car dealerships also speeds up staff tasks like compiling information, writing proposals, and flagging promising deals from buyer interests. Facing a shrinking labour pool, Honda is betting the setup lifts dealership sales efficiency.
Rodland Real Estate — Rodland Real Estate has moved every agent onto Roro, a spoken-command AI real estate platform devised by founder Tim Rodland. Asked aloud, Roro returns median prices, sales volumes, price-per-square-foot and pricing shifts across Old Fort Bay, rebuilt daily from Bahamas Real Estate Association feeds. The brokerage AI system also opens client files, drafts compliance paperwork and handles diaries and correspondence; nobody outside the firm may sign in.
Shinei Shoji — Shinei Shoji, a Japanese wholesale distributor, cut weekly workloads from 52 to 12 hours after a two-month pilot of the JAPAN AI platform—an example of AI agents for wholesale distribution. Staff built 62 custom agents for order emails, document checks and sales reporting, shrinking one report task from 8 hours to 10 minutes. The JAPAN AI platform rollout now covers all 100 employees.
Safaricom — Agentic AI customer service now powers dozens of workflows at Safaricom, letting users finish tasks on digital platforms rather than queuing at a shop or calling in. Dilip Pal, the firm's finance and innovation chief, told Nairobi's De{c0}dE 4.0 summit that cutting friction across multi-step journeys was the goal. Fraud checks, tailored engagement and energy management round out its wider AI push in Kenya.
Clearwater USVI Realty and Property Management — Clearwater USVI Realty and Property Management now claims a US Virgin Islands first: 24/7 AI sales agents in real estate, live on its website to greet visitors day and night. Instead of waiting for office hours, buyers chat with the bot, which records their needs and forwards full transcripts so human staff can pick up the conversation. The Christiansted firm paired this always-on client-intake bot with a fresh property app on the Apple Store, extending its reach across the St. Croix market.
Mauritius Sotheby's International Realty — Mauritius Sotheby's International Realty spent two years building MaxAI, the Mauritian market's first AI real estate agent and AI property assistant. Buyers describe what they want in plain language to surface listings, while staff get automated draft emails and follow-ups.
Arcor — Over 250,000 kiosks, grocery stores and self-service shops in Argentina gain AI-driven inventory management for small retailers via Tokin, Arcor's new AI-powered B2B retail platform. It scans 600-plus variables per outlet to customise stock suggestions, matches shops with Arcor's goods and 80-plus partner brands, then links them to official distributors for restocking.
Raiffeisen Immobilien NÖ/Wien/Burgenland — Raiffeisen Immobilien NÖ/Wien/Burgenland built Ri.Ki, an AI assistant for estate agents, with iteratec: piloted from summer 2025, tested internally by November. Housed in Microsoft Teams, it fetches live records to answer staff on portfolios, energy certificates and tenancy law — part of a wider shift toward AI-assisted real estate services in Austria.
Ooredoo Qatar — Retail shoppers in Qatar gained access to Ooredoo's digital assistant Obot on 26 November 2025, now paired with an avatar feature powered by GPT Real-Time Voice and Azure AI Services. Deployed across X, Facebook and WhatsApp, this AI customer service tool leans on Dynamics 365 Omnichannel so each interaction feels tailored to the individual user. A tool once limited to scripted replies has grown into a fuller virtual assistant, able to converse through text, spoken voice or a lifelike avatar.
Motorpoint — Lily, Motorpoint's AI sales agent for car dealers, closed 350 UK used-car deals in six months to September 2025, pursuing shoppers abandoned by staff. Motorpoint links this AI-assisted car sales push to profits climbing 80% to £2.7 million and turnover climbing 15% to £647.7 million.
Divar — Iran's classifieds platform Divar now lets دستیار فروش, its AI sales assistant, field buyer calls and haggle over prices on sellers' behalf. Within limits sellers define upfront, the tool chats or talks around the clock and fixes viewing appointments without human input. Each national ID unlocks one free run; further use costs ۴۹ هزار تومان on ordinary listings and ۶۹ هزار تومان on job ads under this automated bargaining tool.
Globe Business — A new AI in telecom customer service deployment has cut Globe Business's overall support workload by 34%. The enterprise arm of Globe Telecom built the rollout on Salesforce's Service Cloud, Sales Cloud and Einstein for Service, plus a pilot of Agentforce, sorting incoming emails by intent and directing them straight to the right team. Field staff can now pull up customer records instantly rather than searching manually, and intent-based case classification reaches 80% accuracy on Globe Business's own figures. In the Philippines, this telecom AI customer support model targets sharper personalisation throughout the client relationship.
Metrovacesa — Metrovacesa built MiA, an AI agent for real estate sales, atop Google Cloud's Vertex AI. Response times fell 38%, and early contact handling drove a potential €90.7 million pipeline, with 77% of buyers scoring the multilingual assistant 8.75 out of 10. This AI-powered property sales assistant now runs Spain's top homebuilder's qualification and commercial workflows.
ÍNTEGRO (Corporación Aicsa) — Corporación Aicsa's ÍNTEGRO arm introduced Alana, Guatemala's first AI real estate agent, fielding calls, arranging bookings and messaging buyers across commercial, residential and hospitality deals. This AI-powered property agent launch drew 9.8 million views plus almost 15 mil followers on @alanasliving within two months, a figure the company reports itself.
Tele2 Latvia — Tele2 Latvia has launched Dzirkstele, an AI shopping assistant on its e-shop, to guide customers choosing mobile and home internet plans and to draft product descriptions. Built for AI customer service in telecoms, it works across multiple languages, reading input for tone and intent to generate tailored answers instead of scripted replies. This conversational AI assistant favours context-aware responses.
Proximus NXT — Proximus NXT, the B2B arm of Proximus, has rolled out uman, an AI platform for AI in B2B sales that supports account managers handling complex deals. The system pulls together relevant documents, product details and client background so meeting preparation takes minutes instead of hours, while also helping with follow-up work. Using AI-assisted account management, Proximus NXT reports the tool cuts preparation workload substantially and boosts cross-selling by making it simpler for staff to pitch products outside their usual range.
Porta da Frente Christie's — Porta da Frente Christie's, a Portuguese real estate brokerage, deployed eSelf AI's conversational agent to field property enquiries and run virtual tours at all hours, across a portfolio of 5,000 properties. This use of AI in real estate produced $100 million in sales from leads generated by the AI-driven property inquiry system, the firm reports, a figure that is the company's own and has not been independently audited.
Fisher & Paykel — Fisher & Paykel's Salesforce AI agent resolves 70 per cent of product troubleshooting queries, versus 30 per cent for previous chatbots, since the New Zealand home appliance maker launched AI customer service in February. An agentic AI support bot briefs technicians with customer profile summaries and suggested parts before home repair visits.
Sellers who use AI well report win-rate gains of over 30%. Given that selling typically fills only about a quarter of sellers' time, Bain's 2026 B2B Growth Agenda measured AI 'knowledge assistants' adding 25-35% to effective selling time, grounding the earlier 'AI could double selling time' hope.
AI saves sellers an average 4.8 hours a week, yet 72% of sales organisations fail to put that time into high-value selling, a Gartner survey of 210 chief sales officers and senior leaders (early 2026) found. The issue is now managing the freed time, not whether sales AI helps.
69% of B2B buyers turn to a salesperson to validate AI-generated insights, Gartner found in May 2026. Even as AI research spreads, buyers still lean on people for trust and quality checks.
Sales organisations giving sellers AI-enabled 'next best actions' are 2.6x likelier to grow commercially (Gartner, May 2026). It forecasts that 95% of seller research workflows will start with AI by 2027, from under 20% in 2024.
Sellers who partner effectively with AI tools are 3.7 times likelier to hit quota than those who do not. Gartner surveyed 1,026 B2B sellers from January to March 2024.
Leading adopters of sales AI achieved about 2x the AI-attributed revenue growth and 1.8x the cost efficiency of laggards in Bain's B2B Growth Agenda 2026 (1,125 sales and marketing leaders, 40 countries). AI 'knowledge assistants' added 25-35% to effective selling time, grounding 'double the selling time' hopes.
By 2028 AI sales agents will outnumber human sellers 10 to 1, yet under 40% of sellers will say AI agents raised their productivity (Gartner forecast, 18 Nov 2025). That tempers AI-SDR hype.
Agendor — Agendor's Panorama of AI Use in B2B Sales in Brazil surveyed 1,031 professionals and found 82% expect AI in B2B sales to lift their output. Over half, 52.38%, predict machines will take over routine operational work within two years, whereas 30.26% instead picture the technology as a strategic partner guiding decisions rather than just a tool for machines taking over routine operational work.
Billcom — Billcom, a Tokyo-based PR agency, has rolled out 36 AI agents across every department as part of its shift to becoming an 'AI Native PR Company', marking a notable case of AI in public relations. The agents draw on 23 years of accumulated PR expertise and proprietary media, trend and analytics data built up since the firm's founding. Automated systems now handle routine execution tasks, freeing staff to concentrate on client relationships and strategic decisions, illustrating how AI-driven PR automation can reshape agency workflows.
Mercari — Mercari's Corporate Communication Division in Japan says its CC Agent for AI in corporate communications shrank in-house coordination to two business days from one week. The AI communications agent bundles 20-plus skills for setting strategy, drafting material and tracking press coverage, with writing and quality rules as safeguards.
The UK Government Communication Service's in-house generative-AI toolset, 'Assist', reaches over 4,500 communicators in more than 210 organisations, above 70 per cent of potential users. GCS reports about three hours saved per user weekly, some GBP 5.5 million in projected annual efficiencies and 98 per cent of surveyed users finding it useful (Net Promoter Score 62). The first cross-government generative-AI tool to go 'live', it won a 2025 National AI Award.
CDAC Network — Sudan's information war now has an algorithmic watchdog: CDAC Network and Valent trained an AI disinformation detection system exposing fake networks and coordinated online manipulation. Results reach humanitarian responders in Arabic and English, sharpening decisions amid Sudan's conflict via this AI-driven narrative manipulation monitoring effort.
Burson México — Decipher, a new tool from Burson México, ranks as the region's first cognitive AI system for message credibility prediction. It scans structured and unstructured data to spot reputational threats and gauge how far a message might spread before release, pushing communicators from firefighting into predictive communications strategy.
dentsu México — Playground, an AI advertising platform first built by Merkle LATAM in Colombia, has moved from pilot to daily use across dentsu México's agency network in Mexico. Staff there now rely on it to automate reporting and speed up campaign optimisation, drawing on eye tracking to assess creative work, modules for search optimisation, and a bespoke chatbot trained on 60,000 real consumer conversations. That scale of training data grounds the chatbot's responses in actual buyer language rather than scripted assumptions.
Roughly 60% of PR, media and communications professionals name misinformation as their top AI concern, and about 60% admit using generative-AI tools without approval ('shadow AI'). It leaves a trust and governance gap in the teams that handle AI disclosure.
PR's AI use is maturing while headcount adoption plateaus: 75% of PR professionals use at least one paid AI tool (from 57%), 51% have a formal AI-use policy (from 21% in 2024) and 43% get AI training at work (from 21%). Yet only 12% of AI users use AI agents.
76% of PR professionals say they use generative AI, flat on a year earlier. Uptake has plateaued after years of fast growth.
60% of communications professionals think AI will benefit PR. 28% expect a negative impact.
Social media content leads AI in PR (43%). PR AI then serves research and analytics, and press materials, at 36% each.
Sebrae — Sebrae has built an AI business-planning assistant on Google's Gemini into PNBOX+, free on the Sebrae Portal since 23 September. The conversational agent guides Brazil's entrepreneurs through a plan, clarifies management ideas and poses questions; technical director Bruno Quick stressed that the tool backs owners rather than drafting plans. PNBOX+ logged over 167,000 consultations, almost 18,000 active users and more than 200,000 plans in 2026.
Reunert Group — Anthonie de Beer, group CEO of JSE-listed Reunert, has pushed AI in corporate business units across 17 divisions, leaning on Anthropic's Claude, Google's Gemini and Microsoft's Copilot. His goal: grow revenue, not cut jobs — staff now move faster through decisions, research and meeting prep. Engineering teams have added AI-powered analytics too, an AI adoption across company divisions finishing jobs far quicker.
Kirin Holdings Company, Limited — CoreMate, an AI executive built by Kirin Holdings, now sits on the Kirin Group Executive Committee to guide corporate strategy. 12 distinct AI personas power it, drawing on 10 years' worth of board records, internal files and outside sources, then producing the issues and viewpoints that emerge when those personas debate each other during live meetings. Operating-company executive committees and Kirin's board of directors are next in line for this AI boardroom assistant.
Moderna rolled out OpenAI's ChatGPT Enterprise company-wide and let staff build AI assistants; in about two months they created over 750 custom GPTs across legal, research, manufacturing and commercial functions. Legal reached 100 per cent adoption, the highest of any department, the company says.
Just 20% of companies, the 'AI leaders', capture 74% of AI's economic value through growth and reinvention, not productivity alone. It is a stark value gap.
Although roughly 88-90% of organisations report using AI, only about 37% say it has produced enterprise-level EBIT impact, roughly unchanged on a year earlier, McKinsey's State of AI: Global Survey 2026 finds. That is the adoption-to-value gap in one number.
About 95% of enterprise generative-AI pilots showed no measurable P&L impact and only 5% captured significant value, an MIT study reported. Caution: it is self-labelled 'preliminary findings', not peer-reviewed, its methodology has been publicly questioned, and its authors at MIT Media Lab's Project NANDA build AI-agent infrastructure, a structural conflict of interest.
PwC Channel Islands — PwC's Channel Islands CEO Survey 2026, 'Catalysing competitiveness in the age of AI', polled 45 chief executives across Guernsey and Jersey. Channel Islands AI adoption looks shallow: 58% doubt the firm can shift quickly enough to match technology, and merely 8% see revenue rise and costs fall together. On AI maturity, just 15% have sufficiently many of the seven foundations for success; 80% deem innovation indispensable, yet only 9% claim solid innovation footings.
88% of surveyed organisations used AI in 2025, up from 78% in 2024, the 2026 AI Index reports. 70% used generative AI in at least one business function.
56% of chief executives saw no revenue rise or cost fall from AI investments over the past 12 months. PwC's 29th Global CEO Survey (January 2026) polled 4,454 CEOs in 95 countries.
About 28% of respondents at AI-using organisations report CEO-level oversight of AI governance, and roughly 17% board-level, McKinsey's State of AI: Global Survey 2026 (published August 2026) finds. Top-level oversight has barely moved in a year.
76% of organisations now have a Chief AI Officer, up from 26% a year earlier. The fastest-growing C-suite role shows AI governance climbing the org chart.
Only 34% of organisations use AI to transform their work deeply, with 'talent readiness' the weakest capability at 20%, down 2 points in a year. The bottleneck is depth of transformation rather than adoption.
PwC Bermuda — PwC's Caribbean Corporate Governance Survey 2026 - Bermuda Edition, June 2026, has 62 per cent of Bermuda directors seeing gains from AI and generative AI spending, yet only 29 per cent credit management with delivering AI and GenAI strategy. AI governance ties cybersecurity as top risk, each primary for 24 per cent; 41 per cent back the board's capacity to oversee AI and GenAI plans, 37 per cent confirm a formal route for AI-related opportunities and risks, 37 per cent report thin board education. Just 7 per cent rate AI expertise top when recruiting, and PwC Bermuda partner Marisa Savage said boardrooms now demand different skills, leaving board AI oversight short.
South Korea
United States
Kazakhstan
United Kingdom
Ecuador
Belgium
India
Russia
Spain
Egypt
Japan
Brazil
Chile
Indonesia
Argentina
Dominican Republic
Vietnam
Taiwan
El Salvador
Singapore
Austria
Kyrgyzstan
Nigeria
Curaçao
Germany
Philippines
Uzbekistan
Italy
Netherlands
Bangladesh
Greece
United Arab Emirates
Macao
Tajikistan
Canada
Paraguay
Malaysia
Morocco
Belarus
South Africa
Mali
Liberia
Senegal
Iraq
Peru
Pakistan
Serbia
Algeria
Oman
Bolivia
France
China
Nepal
Lithuania
Switzerland
Mexico
Jordan
Panama
Turkey
Uganda
Montenegro
Albania
Guatemala
Kenya
Ukraine
Hong Kong
Mayotte
Tanzania
Croatia
Poland
Sri Lanka
Armenia
Rwanda
Honduras
Thailand
Colombia
Andorra
Cambodia
Somalia
Qatar
Ethiopia
Uruguay
Venezuela
New Zealand
Djibouti
Belize
Barbados
Eswatini
Cyprus
Luxembourg
Portugal
Yemen
Sweden
Bahrain
Mauritania
North Macedonia
Ghana
Mauritius
Bahamas
Costa Rica
Turkmenistan
Saudi Arabia
Liechtenstein
Guyana
Czechia
Tunisia
Israel
Gambia
Libya
Georgia
Botswana
Fiji
French Guiana
Zimbabwe
Burkina Faso
Laos
Slovenia
Gibraltar
Mongolia
Togo
New Caledonia
Bermuda
Hungary
Iran
Kuwait
Azerbaijan
Palestine
Ireland
Romania
Australia
Finland
Bulgaria
Cote d'Ivoire
Iceland
Réunion
Malta
Turks and Caicos Islands
Norway
Isle of Man
Lebanon
Moldova
Maldives
Denmark
Madagascar
Latvia
Papua New Guinea
Bosnia and Herzegovina
Trinidad and Tobago
Jamaica
Slovakia
Gabon
Cuba
Faroe Islands
Zambia
Benin
Suriname
Saint Lucia
French Polynesia
Brunei
Jersey
Antigua and Barbuda
Malawi
North Korea
Myanmar
Namibia
Angola
Estonia
DR Congo
Cameroon
Sierra Leone
Equatorial Guinea
Niger
Sudan
Congo
Guinea
Seychelles
Nicaragua
Bhutan
Guam
European Union
Chad
Haiti
Cabo Verde
Puerto Rico
Åland Islands
Guadeloupe
Vatican City
Mozambique
Burundi
Cayman Islands
Guernsey
Vanuatu
Monaco
Martinique
South Sudan
Central African Republic
Syria
Greenland