# Straits Institute for Applied AI

> The training, publishing and academic research body of Straits AI Group and iGroup, based in Singapore. It builds AI education for the people who do the work — profession-specific curriculum, taught and self-paced programmes, and its own applied research.

This file is the full text of the Institute's live record of applied AI. For the
page-by-page index, see https://www.straitsai.institute/llms.txt. The same record is available as
structured data at https://www.straitsai.institute/state-of-applied-ai/evidence.json.

Source: https://www.straitsai.institute/state-of-applied-ai
Attribution: Industrial Research Unit, Straits Institute for Applied AI
Last revised 2026-09-14
1114 entries · 17 industries · 249 countries tracked · 1114 carrying a source link

Each entry is a named organisation putting AI into a real job — with the evidence
and rules forming around it — tagged by kind, industry, region and the role it
changes, and dated. Four kinds: Deployment (a named organisation putting AI to
work — the deployment is the news); Evidence of impact (a measured result, trials
and documented outcomes, including failures); Evidence of adoption (how widely AI
is being taken up); Regulation (laws and official guidance).

# The record, industry by industry

Every entry below is reproduced as recorded, with the source that reported it and
the date. Nothing here is inferred.

## Government & Public Sector

- [Deployment] United States · Defence staff — 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. [militarytimes.com](https://militarytimes.com/industry/techwatch/2026/08/31/the-militarys-chatgpt-is-now-live-via-the-pentagons-genai-platform/), 31 Aug 2026
- [Deployment] United States · HR officer — 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. [thedefensepost.com](https://thedefensepost.com/2026/08/07/us-army-ai-hr-support/), 7 Aug 2026
- [Deployment] United States · Logistics planner — 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. [federalnewsnetwork.com](https://federalnewsnetwork.com/army/2026/08/dod-building-ai-tools-to-predict-mitigate-risks-to-sustainment-supply-chain/), Aug 2026 (testing to Mar 2027)
- [Deployment] United States · Intelligence analyst — 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. [reason.com](https://reason.com/2026/08/28/the-cia-is-making-ai-generated-reports-for-local-police/), 28 Aug 2026
- [Deployment] United States · Customer-service agent — 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. [develop.fedscoop.com](https://develop.fedscoop.com/va-launches-ai-helpdesk-chatbot/), 28 Aug 2026
- [Deployment] United States · Emergency dispatcher — 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. [newsweek.com](https://newsweek.com/new-orleans-ai-911-calls-other-cities-12303090), 6 Aug 2026
- [Deployment] United Kingdom · Police officer — 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. [biometricupdate.com](https://biometricupdate.com/202608/british-transport-police-extends-live-facial-recognition-trial-into-underground-stations), 11 Aug 2026
- [Deployment] United Kingdom · Police officer — 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. [biometricupdate.com](https://biometricupdate.com/202608/uk-police-widen-live-facial-recognition-deployments-despite-opposition), 19–21 Aug 2026
- [Deployment] United Arab Emirates · Civil servant — 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. [aiagentstore.ai](https://aiagentstore.ai/ai-agent-news/daily/2026-08-18), 18 Aug 2026
- [Deployment] United States · Logistics planner — 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. [aiagentstore.ai](https://aiagentstore.ai/ai-agent-news/daily/2026-08-06), 6 Aug 2026
- [Deployment] United Kingdom · Call handler — 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. [publictechnology.net](https://www.publictechnology.net/2026/08/21/public-order-justice-and-rights/home-office-targets-ai-generated-savings-with-police-non-emergency-phone-line/), 2026-08-21
- [Deployment] Greece · Tax auditor — 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. [in.gr](https://www.in.gr/2026/08/24/english-edition/tax-authoritys-new-structure-and-ai-units-come-into-force/), 2026-08-24
- [Deployment] Malaysia · Citizen-services officer — 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. [digital.gov.my](https://www.digital.gov.my/en-GB/siaran/MyGOV-Malaysia-Melangkah-Ke-Era-Baharu-Dengan-Agentic-AI), 2026-08-10
- [Deployment] Hong Kong · Public-service officer — 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. [news.gov.hk](https://www.news.gov.hk/eng/2026/08/20260817/20260817_150017_056.html), 2026-08-17
- [Deployment] China · Community officer — 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. [en.people.cn](http://en.people.cn/n3/2026/0825/c90000-20492068.html), 2026-08-25
- [Deployment] Japan · Civil servant — 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. [japantoday.com](https://japantoday.com/category/tech/japan-gov't-launches-in-house-developed-ai-to-assist-civil-servants), 2026-08-22
- [Deployment] Japan · Municipal officer — 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. [itbusinesstoday.com](https://itbusinesstoday.com/tech/business-tech/soka-city-brings-on-premise-generative-ai-into-government-operations/), 2026-08-28
- [Deployment] Thailand · Citizen-services officer — 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. [nationthailand.com](https://www.nationthailand.com/business/tech/40069488), 2026-08-06
- [Deployment] South Korea · Citizen-services officer — 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. [koreajoongangdaily.com](https://www.koreajoongangdaily.com/business/sk-telecom-kakao-and-kt-chosen-for-public-ai-service-project/12849094), 2026-08-28
- [Deployment] India · Rail operations — 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. [republicworld.com](https://www.republicworld.com/tech/indian-railways-ai-ticket-reservation-system-waitlist-predictions-august-2026-06-09-127495), 2026-08-15
- [Deployment] India · Tax officer — 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. [blog.saginfotech.com](https://blog.saginfotech.com/ap-unveils-ai-ml-based-gst-administration-model-6th-ncm), 2026-08-15
- [Deployment] Pakistan · Tax auditor — 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. [propakistani.pk](https://propakistani.pk/2026/07/26/govt-to-use-ai-to-revive-fbr/), 2026-07-26
- [Deployment] United Arab Emirates · Judge / court officer — 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. [gulfnews.com](https://gulfnews.com/uae/government/uae-launches-worlds-first-fully-integrated-ai-powered-judicial-platform-1.500622338), 2026-07-28
- [Deployment] Egypt · Tax officer — 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. [dailynewsegypt.com](https://www.dailynewsegypt.com/2026/08/29/five-steps-to-obtain-certified-online-clearance-for-real-estate-transaction-tax-payments-finance-minister/), 2026-08-19
- [Deployment] Brazil · Citizen-services officer — 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. [floripamanha.org](https://floripamanha.org/2026/08/florianopolis-e-a-primeira-capital-a-usar-ia-generativa-para-atendimento-ao-publico/), 2026-08-27
- [Deployment] Brazil · Citizen-services officer — 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. [ohoje.com](https://ohoje.com/2026/08/14/rio-verde-registra-65-mil-atendimentos-com-uso-de-inteligencia-artificial/), 2026-08-14
- [Deployment] Chile · Tax call agent — 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. [tvn.cl](https://www.tvn.cl/exponencial/noticias/agentes-ia-servicio-publico-chile), 2026-07-23
- [Deployment] South Africa · Public servant — 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. [timeslive.co.za](https://www.timeslive.co.za/news/2026-08-05-innovation-meets-impact-as-minister-buthelezi-launches-batho-pele-ai-chatbot/), 2026-08-01
- [Deployment] Rwanda · Planning officer — 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. [spaceinafrica.com](https://spaceinafrica.com/2026/08/10/rwanda-and-planet-launch-landmark-national-satellite-data-programme/), 2026-08-10
- [Evidence of adoption] Global — 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. [oecd.org](https://www.oecd.org/content/dam/oecd/en/publications/reports/2024/06/governing-with-artificial-intelligence_f0e316f5/26324bc2-en.pdf), 2023
- [Evidence of adoption] United Kingdom — 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. [nao.org.uk](https://www.nao.org.uk/wp-content/uploads/2024/03/use-of-artificial-intelligence-in-government.pdf), 2024
- [Evidence of impact] United States — 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. [pa.gov](https://www.pa.gov/governor/newsroom/2025-press-releases/-shapiro-administration-leads-the-way-in-ethical-use-of-ai), 2025
- [Deployment] United Kingdom · Civil servant — 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. [gov.uk](https://www.gov.uk/government/news/ground-breaking-use-of-ai-saves-taxpayers-money-and-delivers-greater-government-efficiency), 
- [Evidence of adoption] United States — 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). [github.com](https://github.com/ombegov/2025-Federal-Agency-AI-Use-Case-Inventory), 2024
- [Deployment] United States · Fire / emergency — 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. [route-fifty.com](https://www.route-fifty.com/public-safety/2026/07/what-california-has-learned-ai-enabled-early-wildfire-detection-system/414566/), 2025
- [Regulation] European Union — 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. [digital-strategy.ec.europa.eu](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), 2026
- [Regulation] European Union — 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. [artificialintelligenceact.eu](https://artificialintelligenceact.eu/article/5/), 2025
- [Evidence of adoption] United States — 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. [nextgov.com](https://www.nextgov.com/artificial-intelligence/2026/01/law-enforcement-leading-dhs-use-case-ai/411063/), 2025
- [Deployment] United Kingdom · Police officer — 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. [gov.uk](https://www.gov.uk/government/news/policeai-to-speed-up-investigations-and-fight-crime), 2026
- [Evidence of impact] United Kingdom — 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. [gov.uk](https://www.gov.uk/government/publications/police-use-of-facial-recognition/police-use-of-facial-recognition-factsheet), 2025
- [Regulation] United Kingdom — 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. [biometricupdate.com](https://www.biometricupdate.com/202608/uk-ico-finds-police-facial-recognition-use-mostly-compliant-with-data-regulations), 2026
- [Regulation] United States — 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. [police1.com](https://www.police1.com/artificial-intelligence/conn-officials-pause-statewide-le-usage-of-ai-report-writing-software), 2026
- [Evidence of impact] Global — 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. [jamanetwork.com](https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2774644), 2021
- [Deployment] Global · Fire / emergency — 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. [gov.ca.gov](https://www.gov.ca.gov/2026/07/07/californias-wildfire-defense-blasts-off-governor-newsom-launches-firesat-wildfire-detection-satellites-to-spot-blazes-from-space/), 2026
- [Evidence of adoption] United Kingdom — 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. [gov.uk](https://www.gov.uk/government/news/pm-launches-national-skills-drive-to-unlock-opportunities-for-young-people-in-tech), 2026
- [Evidence of adoption] European Union — 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. [digital-strategy.ec.europa.eu](https://digital-strategy.ec.europa.eu/en/policies/digital-skills), 2026
- [Evidence of adoption] South Korea — 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. [koreaherald.com](https://www.koreaherald.com/article/10612847), 2026
- [Deployment] Brazil · judge — 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. [web.trf3.jus.br](https://web.trf3.jus.br/noticias/Noticiar/ExibirNoticia/446267-cnj-lanca-lia3r-e-amplia-uso-de-inteligencia-artificial), 2026-08-20
- [Deployment] Bolivia · public servant — 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. [urgente.bo](https://www.urgente.bo/noticia/planificacion-lanza-bolivia-el-primer-modelo-de-inteligencia-artificial-del-estado), 2026-08-09
- [Regulation] Brazil — 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. [clickondetroit.com](https://www.clickondetroit.com/news/world/2026/09/02/brazils-electoral-court-sets-new-standards-for-ai-deepfakes-after-case-involving-sen-bolsonaro/), 2026-09-01
- [Regulation] Chile — 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. [linaresenlinea.cl](https://www.linaresenlinea.cl/2026/08/31/nueva-red-impulsara-verificacion-independiente-del-uso-de-inteligencia-artificial-en-organismos-publicos/), 2026-08-31
- [Evidence of adoption] Chile — 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. [latercera.com](https://www.latercera.com/nacional/noticia/ministerio-de-ciencia-fija-seis-ejes-para-2026-2030-y-prepara-nuevo-proyecto-de-ley-de-inteligencia-artificial/), 2026-08-13
- [Evidence of adoption] Bahamas — 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. [tribune242.com](https://www.tribune242.com/news/2026/aug/13/govt-to-invest-3m-into-ai-skills-training/), 2026-08-13
- [Evidence of adoption] Saint Kitts and Nevis — 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. [eccb-centralbank.org](https://www.eccb-centralbank.org/news/eccu-generative-ai-and-python-summer-camp-concludes-with-regional-showcase-of-chatbots-and-multimedia-projects), 2026-08-28
- [Deployment] Mexico · police officer — 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. [infobae.com](https://www.infobae.com/mexico/2026/08/08/san-luis-potosi-despliega-perros-robot-con-inteligencia-artificial-para-reforzar-la-seguridad-publica/), 2026-08-06
- [Deployment] Honduras · municipal services officer — 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. [hondudiario.com](https://www.hondudiario.com/2026/08/18/alcaldia-crea-amdcia-y-lanza-mdc-en-tus-manos-el-primer-agente-municipal-de-inteligencia-artificial-de-honduras/), 2026-08-18
- [Regulation] Panama — 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. [gmalabs.com](https://www.gmalabs.com/post/panama-national-ai-strategy-approved-resolution-89-26), 2026-08-12
- [Deployment] United States · government customer service agent — 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. [statescoop.com](https://statescoop.com/after-successful-dmv-pilot-nebraska-expands-tyler-technologies-ai-tools-statewide/), 2026-09-01
- [Deployment] United States · civil servant — 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. [aol.com](https://www.aol.com/articles/california-state-workers-access-ai-181624000.html), 2026-08-19
- [Deployment] Canada · traffic engineer — 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. [globalnews.ca](https://globalnews.ca/news/12039926/city-of-calgary-ai-high-risk-intersections-for-pedestrians/), 2026-08-28
- [Deployment] Canada · police officer — 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. [ca.news.yahoo.com](https://ca.news.yahoo.com/constant-eyes-toronto-police-ai-080000288.html), 2026-08-31
- [Deployment] China · Police crime-prevention officer — 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. [global.chinadaily.com.cn](http://global.chinadaily.com.cn/a/202609/06/WS6a9d19f7e4b06d4aa055c946.html), 2026-09-06
- [Evidence of impact] Hong Kong — 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. [scmp.com](https://www.scmp.com/news/hong-kong/society/article/3364317/hong-kong-halves-environmental-impact-review-times-ai-efficiency-drive), 2026-08-17
- [Evidence of adoption] Malaysia — 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. [thestar.com.my](https://www.thestar.com.my/news/nation/2026/09/04/free-ai-skills-for-100000-youths), 2026-09-04
- [Deployment] Pakistan · Intelligence analyst — 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. [app.com.pk](https://www.app.com.pk/domestic/ai-enabled-intelligence-system-to-strengthen-punjabs-security-responsecm/), 2026-09-05
- [Evidence of adoption] Pakistan — 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. [phoneworld.com.pk](https://www.phoneworld.com.pk/hec-concludes-artificial-intelligence-training-for-two-cohorts-of-around-400-faculty-members/), 2026-09-07
- [Deployment] Japan · Local government officer — 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. [govtechtokyo.or.jp](https://www.govtechtokyo.or.jp/news/2026/04/09/4918/), 2026-04-09
- [Regulation] Malaysia — 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. [digital.gov.my](https://www.digital.gov.my/en-GB/siaran/AI-Malaysia-Pemacu-Utama-Menuju-Negara-AI-2030), 2026-07-28
- [Evidence of adoption] Malaysia — 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. [talentcorp.com.my](https://www.talentcorp.com.my/resources/press-releases/talentcorp-drives-national-ai-readiness-as-rm110-million-jelajah-ai-mymahir-expands-nationwide/), 2026-01-17
- [Regulation] Philippines — 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. [quisumbingtorres.com](https://www.quisumbingtorres.com/en/alerts/2026/07/framework-for-ethical-and-responsible-ai-use-in-govt-issued), 2026-06-09
- [Evidence of adoption] Philippines — 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. [tribune.net.ph](https://tribune.net.ph/2026/01/09/deped-rolls-out-project-agapai), 2026-01-09
- [Deployment] Sweden · Public-sector caseworker — 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. [skatteverket.se](https://www.skatteverket.se/omoss/pressochmedia/nyheter/2026/nyheter/skatteverketbyggergrundenforaianvandninginomoffentligsektor.5.3129d65419ef1497c0610c7.html), 2026-07-01
- [Deployment] Romania · Citizen-services call handler — 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. [connect.ro](https://www.connect.ro/2026/04/21/inteligenta-artificiala-primaria-sector-3/), 2026-04
- [Deployment] Peru · Judge — 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. [elperuano.pe](https://www.elperuano.pe/noticia/286434-poder-judicial-implementa-herramienta-ia-chatbot-curia-en-tres-cortes-superiores), 2026-01-04
- [Deployment] Mexico · Judge — 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. [elimparcial.com](https://www.elimparcial.com/mexico/2026/04/30/por-primera-vez-en-mexico-una-inteligencia-artificial-ayudo-a-redactar-dos-sentencias-judiciales-en-queretaro-y-el-experimento-abrio-un-debate-urgente-sobre-que-tanto-puede-acelerar-la-justicia-sin-poner-en-manos-de-una-maquina-la-decision-final-de-jueces-y-juezas/), 2026-04-29
- [Deployment] Argentina · Court clerk — 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. [justicia.ar](https://www.justicia.ar/novedades/4757/2026/06/el-poder-judicial-incorpora-una-herramienta-de-anonimizacion-con-inteligencia-artificial), 2026-06-23
- [Deployment] Turkey · Judge — 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. [bianet.org](https://bianet.org/haber/hakim-ve-savcilar-icin-orgut-tahmini-yapan-yapay-zeka-sistemi-322362), 2026-06-26
- [Deployment] Israel · Tax auditor — 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. [calcalistech.com](https://www.calcalistech.com/ctechnews/article/ygbeaanqw), 2026-06
- [Evidence of adoption] Turkey — 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. [aa.com.tr](https://www.aa.com.tr/tr/gundem/yapay-zeka-destekli-mebi-platformu-6-milyona-yakin-kullaniciya-ulasti/4006611), 2026-07
- [Deployment] Canada · Government software developer — 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. [medicinehatnews.com](https://medicinehatnews.com/news/local-news/2026/07/07/province-boasting-achievements-in-ai-for-government-tech/), 2026-07-07
- [Deployment] United States · State civil servant — 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. [statescoop.com](https://statescoop.com/massachusetts-chatgpt-ai-executive-branch/), 2026-02
- [Deployment] United States · Judge — 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. [calmatters.org](https://calmatters.org/economy/technology/2026/05/ai-los-angeles-riverside-courts/), 2026-02
- [Deployment] Kenya · Court transcriber — 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. [connectingafrica.com](https://www.connectingafrica.com/ai/kenya-s-judiciary-embraces-ai-digitization), 2026-05
- [Deployment] Ghana · Customs officer — 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. [ecofinagency.com](https://www.ecofinagency.com/news/3004-55175-ghana-navigates-trade-friction-as-ai-reshapes-customs-enforcement), 2026-02-01
- [Deployment] South Africa · Tax auditor — 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. [techcabal.com](https://techcabal.com/2026/07/27/ai-south-africas-taxman/), 2026-07-01
- [Evidence of adoption] Philippines — 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. [sunstar.com.ph](https://www.sunstar.com.ph/cebu/peza-opens-ai-academy-in-cebu), 2026-01-16
- [Evidence of adoption] Sweden — 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. [cioanalytics.se](https://www.cioanalytics.se/artikel/2026/09/nastan-var-tionde-riksdagsmotion-ai-behandlad-utan-att-det-redovisas/), 2026-09-04
- [Deployment] Switzerland · Member of parliament — 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. [radiocentral.ch](https://www.radiocentral.ch/news/schweiz/das-schweizer-parlament-testet-ki-assistenz-pia-165360780), 2026-09-04
- [Regulation] Russia — 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. [united24media.com](https://united24media.com/world/russia-passes-law-requiring-domestic-artificial-intelligence-to-align-with-traditional-values-20592), 2026-09-01
- [Deployment] United Arab Emirates · Government policy adviser — 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. [tbreak.com](https://tbreak.com/uae-cabinet-ai-advisor/), 2026-09-03
- [Deployment] United Arab Emirates · Police administrator — 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. [emirates247.com](https://www.emirates247.com/uae/dubai-police-deploy-36-robots-as-ai-takes-on-bigger-role-across-operations/5297), 2026-09-06
- [Deployment] United Arab Emirates · Property registrar — 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. [mediaoffice.ae](https://www.mediaoffice.ae/en/news/2026/september/03-09/dubai-land-department-launches-initial-registration-to-boost-real-estate-efficiency), 2026-09-03
- [Deployment] United Arab Emirates · Immigration officer — 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. [emirates247.com](https://www.emirates247.com/uae/click-and-travel-dubai-launches-mobile-passenger-clearance-in-as-little-as-12-seconds/5314), 2026-09-07
- [Deployment] Jordan · Government spokesperson — 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. [skynewsarabia.com](https://www.skynewsarabia.com/varieties/1890116), 2026-09-03
- [Deployment] Oman · Government statistician — 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. [muscatdaily.com](https://www.muscatdaily.com/2026/09/07/ncsi-unveils-ai-powered-household-income-expenditure-survey/), 2026-09-07
- [Deployment] Iraq · Cybersecurity officer — 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. [964media.com](https://964media.com/715095/), 2026-09-08
- [Regulation] Turkey — 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. [en.milliyet.com.tr](https://en.milliyet.com.tr/news/turkiye-officially-launches-2026-2030-ai-action-plan-7644115), 2026-08-19
- [Deployment] Switzerland · Public administration officer — 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. [zh.ch](https://www.zh.ch/de/news-uebersicht/medienmitteilungen/2026/07/potenzial-von-kuenstlicher-intelligenz-in-der-verwaltung-sicher-nutzen.html), 2026-07-09
- [Deployment] Canada · 311 customer service agent — 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. [uk.news.yahoo.com](https://uk.news.yahoo.com/toronto-rolling-ai-chatbot-help-080000890.html), 2026-06-17
- [Deployment] Philippines · Civil servant — 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. [techpilipinas.com](https://techpilipinas.com/dict-google-cloud-ink-multi-year-ai-cybersecurity-deal), 2026-06-22
- [Deployment] Germany · Municipal administration officer — 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. [ruesselsheim.de](https://www.ruesselsheim.de/presse/2026-07-22-Mit-kuenstlicher-Intelligenz-zu-effizienteren-Verwaltungsprozessen), 2026-07-22
- [Deployment] Uzbekistan · Tax officer — 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. [uz.sputniknews.ru](https://uz.sputniknews.ru/20260804/nalogovyy-komitet-uzbekistana-zapustil-platformu-ai-soliq-na-baze-ii-59485811.html), 2026-08-04
- [Deployment] Uzbekistan · Customs officer — 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. [spot.uz](https://www.spot.uz/ru/2026/07/01/airo/), 2026-07-01
- [Deployment] Japan · Local government officer — 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. [allganize.ai](https://www.allganize.ai/ja/ja-news/20260730), 2026-07-30
- [Deployment] Pakistan · Anti-corruption investigator — 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. [dawn.com](https://www.dawn.com/news/2010341), 2026-06-24
- [Deployment] Australia · Police officer — 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. [abc.net.au](https://www.abc.net.au/news/2026-08-11/ai-police-face-screening-trial-sparks-privacy-concern/107009644), 2026-08-11
- [Deployment] India · Court registry official — 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. [republicworld.com](https://www.republicworld.com/tech/ai-comes-to-indias-supreme-court-top-court-launches-su-sahay-chatbot-and-unified-digital-case-system-2026-05-11-123861), 2026-05-11
- [Deployment] Kazakhstan · Judge — 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. [ru.elordainfo.kz](https://ru.elordainfo.kz/raznoe-ru/verhovnyy-sud-vnedryaet-ii-analitiku-vo-vseh-sudah-kazahstana), 2026-05-26
- [Deployment] Bangladesh · Traffic police officer — 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. [thesun.my](https://thesun.my/news/world-news/asia/bangladesh-puts-ai-in-driving-seat-to-tackle-terrible-traffic/), 2026-05-24
- [Deployment] Mongolia · Government planning officer — 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. [un-page.org](https://www.un-page.org/news/ai-tuslac-how-mongolia-is-using-ai-to-strengthen-development-planning/), 2026-04-20
- [Deployment] South Korea · Tax counsellor — 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. [korea.kr](https://www.korea.kr/news/policyNewsView.do?newsId=156757496), 2026-04-23
- [Regulation] New Zealand — 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. [scoop.co.nz](https://www.scoop.co.nz/stories/PA2605/S00279/improving-welfare-system-efficiency.htm), 2026-05-29
- [Regulation] Pakistan — 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. [ljcp.gov.pk](https://www.ljcp.gov.pk/naig), 2026-04-29
- [Regulation] Philippines — 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. [philstar.com](https://www.philstar.com/headlines/2026/03/21/2515797/sc-outlines-ai-use-court-operations), 2026-03-21
- [Regulation] South Korea — 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. [cooley.com](https://www.cooley.com/news/insight/2026/2026-01-27-south-koreas-ai-basic-act-overview-and-key-takeaways), 2026-01-27
- [Regulation] Taiwan — 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. [techpolicy.press](https://www.techpolicy.press/taiwans-ai-basic-act-can-be-a-model-for-asia/), 2026-02-25
- [Deployment] Austria · Civil servant — 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. [bundeskanzleramt.gv.at](https://www.bundeskanzleramt.gv.at/bundeskanzleramt/nachrichten-der-bundesregierung/2026/07/proell-public-ai-launcht-govgpt-fuer-die-bundesverwaltung.html), 2026-07-20
- [Regulation] Estonia — 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. [eesti.ai](https://eesti.ai/uudised/eestist-saab-esimene-riik-maailmas-mis-loob-ai-agentidele-digitaalse-identiteedi), 2026-06-17
- [Deployment] Ireland · Passport officer — 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. [rte.ie](https://www.rte.ie/news/2026/0120/1553986-artificial-intelligence-committee/), 2026-01-20
- [Deployment] Iraq · Government planner — 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. [964media.com](https://964media.com/678341/), 2026-05-04
- [Deployment] Argentina · Prison officer — 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. [puebloregional.com.ar](https://puebloregional.com.ar/monitoreo-con-inteligencia-artificial-reconocimiento-facial-e-identificacion-de-voz-para-controlar-llamadas-son-las-nuevas-incorporaciones-de-la-carcel-de-recreo/), 2026-09-03
- [Regulation] Costa Rica — 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. [diarioextra.com](https://www.diarioextra.com/noticia/icd-prohibe-uso-de-inteligencia-artificial-en-la-institucion/), 2026-09-08
- [Deployment] Dominican Republic · Customs inspector — 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. [eldinero.com.do](https://eldinero.com.do/377967/aduanas-incorporara-inteligencia-artificial-al-analisis-de-imagenes-de-rayos-x/), 2026-08-28
- [Regulation] Argentina — 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. [gentedesalta.com](https://www.gentedesalta.com/policiales/los-fiscales-podran-usar-inteligencia-artificial-pero-subir-expedientes-chatgpt-n11619), 2026-09-07
- [Regulation] Argentina — 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. [infobae.com](https://www.infobae.com/politica/2026/08/28/promulgan-ley-anti-deepfake-habra-multas-y-dias-de-arresto-en-chaco-por-casos-de-violencia-digital-con-ia/), 2026-08-28
- [Deployment] South Africa · Investment promotion officer — 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. [capetownetc.com](https://www.capetownetc.com/news/cape-town-launches-ai-investment-guide-maya/), 2026-08-10
- [Evidence of impact] Zimbabwe — 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. [allafrica.com](https://allafrica.com/stories/202608140140.html), 2026-08-14
- [Deployment] Finland · Social worker — 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. [sttinfo.fi](https://www.sttinfo.fi/tiedote/72181535/varsinais-suomen-ja-satakunnan-hyvinvointialueet-laajentavat-tekoalykirjausten-kayttoa-sosiaalihuollossa?publisherId=69820130&lang=fi), 2026-07-06
- [Deployment] Portugal · Municipal planning and licensing officer — 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. [maisalgarve.pt](https://maisalgarve.pt/2026/07/03/portimao-coloca-a-inteligencia-artificial-ao-servico-do-urbanismo-e-agiliza-o-licenciamento/), 2026-07-03
- [Deployment] Bulgaria · Tax officer — 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. [press.azbuki.bg](https://press.azbuki.bg/en/news/novini-2026/nap-vavezhda-balgarskiya-izkustven-intelekt-na-insait/), 2026-03-19
- [Deployment] Greece · Traffic police officer — 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. [zougla.gr](https://www.zougla.gr/greece/psifiaka-erchontai-ta-prostima-apo-tis-kameres-technitis-noimosynis/), 2026-03-30
- [Deployment] Netherlands · Civil servant (central government) — 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. [computable.nl](https://www.computable.nl/2026/06/10/vlam-in-de-ai-pan-bij-rijksambtenaren/), 2026-06-10
- [Deployment] France · France services adviser — 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. [banquedesterritoires.fr](https://www.banquedesterritoires.fr/avec-le-plan-notre-ia-letat-accelere-sur-le-deploiement-des-outils-et-la-gouvernance), 2026-06-16
- [Deployment] Germany · Tax office staff — 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. [finanzverwaltung.nrw.de](https://www.finanzverwaltung.nrw.de/uebersicht-rubrik-aktuelles-und-presse/pressemitteilungen/finanzverwaltung-nordrhein-westfalen-1), 2026-07-27
- [Regulation] Spain — 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. [boe.es](https://www.boe.es/diario_boe/txt.php?id=BOE-A-2026-2205), 2026-01-30
- [Deployment] Luxembourg · Civil servant — 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. [gouvernement.lu](https://gouvernement.lu/fr/actualites/toutes_actualites/communiques/2026/03-mars/04-frieden-ai4lux.html), 2026-03-04
- [Evidence of adoption] Italy — 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. [publicpolicy.net](https://www.publicpolicy.net/ia-sono-59-i-sistemi-in-uso-dallinps-per-migliorare-i-servizi-109550.html), 2026-07-09
- [Deployment] Austria · Federal civil servant (case files) — 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. [ots.at](https://www.ots.at/presseaussendung/OTS_20260319_OTS0089/proell-launcht-public-ai-fuenf-konkrete-ki-anwendungen-fuer-oesterreichs-bundesverwaltung), 2026-03-19
- [Deployment] Mexico · Judicial clerk — 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. [notipress.mx](https://notipress.mx/estado-de-mexico/poder-judicial-edomex-uso-ia-explicar-mil-401-resoluciones-lenguaje-39240), 2026-09-04
- [Deployment] Mexico · Citizen-service officer — 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. [posta.com.mx](https://www.posta.com.mx/tamaulipas/avi-t-el-chat-que-resolvera-tramites-y-dudas-del-gobierno-de-tamaulipas/vl2249386), 2026-08-31
- [Deployment] Peru · Municipal customer-service officer — 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. [andina.pe](https://andina.pe/agencia/noticia-surco-primer-municipio-incorporar-un-asistente-virtual-inteligencia-artificial-1090860.aspx), 2026-09-08
- [Regulation] Peru — 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. [macronorte.pe](https://macronorte.pe/2026/09/10/gobernar-la-ia-o-ser-gobernados-peru-entra-en-una-nueva-etapa-de-regulacion-tecnologica/), 2026-09-10
- [Regulation] Peru — 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. [elperuano.pe](https://elperuano.pe/noticia/304466-sunarp-fija-reglas-para-el-uso-de-ia-registradores-mantendran-responsabilidad-en-las-decisiones), 2026-09-10
- [Deployment] Panama · Citizen-service contact-centre agent — 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. [telemetro.com](https://www.telemetro.com/nacionales/panama-conecta-chatbot-ia-permitira-hacer-reportes-al-311-n6090700), 2026-09-03
- [Deployment] Paraguay · Emergency-response planner — 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. [abc.com.py](https://www.abc.com.py/nacionales/2026/08/29/paraguay-ante-la-amenaza-del-supernino-inteligencia-artificial-proyecta-hasta-100000-familias-damnificadas/), 2026-08-29
- [Deployment] Ghana · Disaster management officer — 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. [gna.org.gh](https://gna.org.gh/2026/05/nadmo-partners-launch-ai-powered-chatbot-to-improve-disaster-risk-communication/), 2026-05-28
- [Deployment] Nigeria · Citizen-services officer — 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. [fmcide.gov.ng](https://fmcide.gov.ng/fmcide-and-meta-launch-govguide-nigeria-an-ai-powered-government-services-chatbot/), 2026-05-21
- [Deployment] Rwanda · Consular officer — 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. [newtimes.co.rw](https://www.newtimes.co.rw/article/33448/news/technology/rwanda-high-commission-in-uk-deploys-ai-to-enhance-service-delivery), 2026-02-20
- [Evidence of impact] Angola — 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. [expansao.co.ao](https://expansao.co.ao/empresas/detalhe/inteligencia-artificial-da-agt-apanha-15-mil-empresas-suspeitas-de-fuga-ao-fisco-73137.html), 2026-07-10
- [Deployment] Ghana · Judge — 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. [adomonline.com](https://www.adomonline.com/judicial-service-to-deploy-ai-assistant-for-all-445-judges-nationwide/), 2026-06-23
- [Deployment] Kenya · County citizen-services officer — 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. [the-star.co.ke](https://www.the-star.co.ke/counties/central/2026-04-20-muranga-rolls-out-ai-powered-app-to-boost-service-delivery), 2026-04-20
- [Evidence of impact] South Africa — 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. [timeslive.co.za](https://www.timeslive.co.za/news/south-africa/2026-07-28-home-affairs-to-expand-digital-visa-system-after-blocking-5500-travellers/), 2026-07-28
- [Deployment] Nigeria · Emergency dispatcher — 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. [newspeddlers.com](https://www.newspeddlers.com/2026/04/lagos-launches-ai-powered-lerms.html), 2026-04-28
- [Deployment] Togo · Tax and market-control inspector — 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. [togofirst.com](https://www.togofirst.com/fr/gouvernance-economique/1303-18453-togo-la-plateforme-digital-market-intelligence-pour-renforcer-le-systeme-de-marquage-des-produits), 2026-03-13
- [Deployment] Burkina Faso · Traffic police officer — 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. [wearetech.africa](https://www.wearetech.africa/fr/fils/actualites/tech/burkina-faso-des-cameras-intelligentes-pour-surveiller-les-infractions-routieres-a-ouagadougou), 2026-07-30
- [Deployment] Cote d'Ivoire · Citizen-services officer — 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. [wearetech.africa](https://www.wearetech.africa/fr/fils/actualites/tech/la-cote-d-ivoire-lance-emy-101-un-chatbot-pour-rapprocher-l-administration-des-citoyens), 2026-03-19
- [Deployment] Djibouti · Weather forecaster — 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. [wearetech.africa](https://www.wearetech.africa/fr/fils/actualites/tech/djibouti-deploie-une-nouvelle-version-de-son-systeme-d-alerte-meteo-fonde-sur-l-ia), 2026-07-22
- [Regulation] Panama — 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. [infobae.com](https://www.infobae.com/panama/2026/06/18/inteligencia-artificial-y-geolocalizacion-se-suman-a-la-alerta-amber-en-panama/), 2026-06-18
- [Deployment] Guatemala · Forensic specialist — 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. [emisorasunidas.com](https://emisorasunidas.com/nacional/2026/07/02/inacif-identifica-victima-herramienta-inteligencia-artificial), 2026-07-02
- [Deployment] Guyana · Traffic police officer — 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. [demerarawaves.com](https://demerarawaves.com/2026/05/22/mobile-ai-system-to-track-intercept-traffic-ticket-evaders/), 2026-05-22
- [Deployment] Belize · Judge — 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. [amandala.com.bz](https://amandala.com.bz/news/high-court-to-deploy-ai-for-quicker-rulings/), 2026-07-16
- [Regulation] Paraguay — 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. [pj.gov.py](https://www.pj.gov.py/notas/29655-corte-suprema-de-justicia-aprobo-politica-para-el-uso-de-sistemas-de-inteligencia-artificial-en-el-poder-judicial), 2026-04-09
- [Regulation] El Salvador — 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. [lanoticiasv.com](https://lanoticiasv.com/sala-de-lo-constitucional-fija-estandares-eticos-y-legales-para-el-uso-de-la-inteligencia-artificial/), 2026-03-16
- [Deployment] El Salvador · Prosecutor — 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. [infobae.com](https://www.infobae.com/el-salvador/2026/04/22/fiscalia-salvadorena-anuncia-sistema-de-inteligencia-artificial-soberana-para-denuncias-judiciales/), 2026-04-21
- [Deployment] Costa Rica · Municipal procurement officer — 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. [delfino.cr](https://delfino.cr/2026/06/municipalidad-de-santa-ana-impulsa-proyecto-de-inteligencia-artificial-para-mejorar-procesos-de-contratacion-administrativa), 2026-06-01
- [Deployment] Colombia · Citizen-service agent — 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. [bogota.gov.co](https://bogota.gov.co/mi-ciudad/gestion-publica/chatico-agente-virtual-de-bogota-se-renueva-con-ia-generativa-2026), 2026-01-29
- [Deployment] Colombia · Police CCTV monitoring operator — 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. [eltiempo.com](https://www.eltiempo.com/colombia/barranquilla/atlantico-implementa-sistema-de-videovigilancia-con-inteligencia-artificial-414-camaras-y-45-000-millones-en-inversion-3529134), 2026-02-02
- [Deployment] Colombia · Early-childhood social worker — 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. [portafolio.co](https://www.portafolio.co/tecnologia/bogota-implementa-sistema-con-inteligencia-artificial-para-anticipar-riesgos-en-la-primera-infancia-492570), 2026-04-23
- [Deployment] Colombia · Citizen-petitions case handler — 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. [mintic.gov.co](https://www.mintic.gov.co/portal/715/w3-article-439551.html), 2026-07-09
- [Deployment] Uruguay · Municipal citizen-service officer — 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. [carmeloportal.com](https://www.carmeloportal.com/124015-florida-prueba-el-futuro-del-tramite-publico-una-intendencia-con-portal-unico-inteligencia-artificial-y-nuevos-desafios-ciudadanos), 2026-07-10
- [Deployment] Argentina · Property-tax inspector — 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. [semanarioactualidad.com.ar](https://semanarioactualidad.com.ar/arba-implemento-m2-un-sistema-digital-con-inteligencia-artificial-para-detectar-construcciones-no-declaradas/), 2026-02-02
- [Evidence of impact] Argentina — 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. [eldiadeescobar.com.ar](https://eldiadeescobar.com.ar/escobar-optimiza-sus-compras-publicas-mediante-inteligencia-artificial/), 2026-01-26
- [Deployment] Argentina · Consumer-protection officer — 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. [noticiasnqn.com.ar](https://www.noticiasnqn.com.ar/noticias/2026/05/07/334056-neuquen-lanzo-un-chatbot-con-inteligencia-artificial-para-asistir-a-consumidores), 2026-05-07
- [Deployment] Argentina · Wildlife warden — 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. [weekend.perfil.com](https://weekend.perfil.com/noticias/pesca/neuquen-suma-drones-con-inteligencia-artificial-para-detectar-pesca-y-caza-furtivas.phtml), 2026-06-16
- [Deployment] Argentina · Citizen-service agent — 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. [inteligenciaargentina.ar](https://inteligenciaargentina.ar/inteligencia-artificial/la-ciudad-de-buenos-aires-lanzo-bax-la-inteligencia-artificial-para-hacer-tramites-y-consultas-desde-una-sola-app), 2026-07-02
- [Deployment] Peru · Benefits programme user-support technician — 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. [andina.pe](https://andina.pe/agencia/noticia-juntos-lanza-maria-chatbot-inteligencia-artificial-orientara-las-24-horas-1074063.aspx), 2026-05-06
- [Evidence of impact] Peru — 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. [revistaeconomia.com](https://www.revistaeconomia.com/sunat-identifica-a-mas-de-34-mil-contribuyentes-por-alquileres-en-plataformas-digitales-y-refuerza-control-con-inteligencia-artificial/), 2026-05-25
- [Deployment] Chile · Customs inspector — 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. [iquiquetv.cl](https://iquiquetv.cl/2026/01/13/entregan-nuevo-camion-escaner-con-ia/), 2026-01-13
- [Deployment] Chile · Product-safety inspector — 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. [cooperativa.cl](https://www.cooperativa.cl/noticias/pais/energia/sec-implementa-inteligencia-artificial-para-detectar-productos/2026-04-17/113826.html), 2026-04-17
- [Deployment] Ecuador · Park ranger — 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. [elcomercio.com](https://www.elcomercio.com/actualidad/ecuador/isla-floreana-ia-vigila-dia-noche-para-traer-vuelta-especies-galapagos/), 2026-06-16
- [Regulation] Ecuador — 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. [eltelegrafo.com.ec](https://www.eltelegrafo.com.ec/noticias/nacionales/210/presidente-noboa-reforma-la-contratacion-publica-e-incorpora-el-uso-de-inteligencia-artificial), 2026-07-30
- [Regulation] Brazil — 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. [convergenciadigital.com.br](https://convergenciadigital.com.br/governo/receita-cria-curador-de-ia-para-cuidar-de-erros-e-vieses-da-inteligencia-artificial/), 2026-03-31
- [Deployment] Brazil · Tax appeals councillor — 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. [migalhas.com.br](https://www.migalhas.com.br/quentes/453045/carf-institui-uso-de-inteligencia-artificial-para-ajudar-conselheiros), 2026-04-01
- [Deployment] Brazil · Reimbursement analyst — 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. [csb.org.br](https://csb.org.br/noticias/inteligencia-artificial-sus-cobrar-dividas-planos-de-saude), 2026-06-02
- [Regulation] Saudi Arabia — 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. [ajel.sa](https://ajel.sa/economy/knc09efrb), 2026-07-14
- [Deployment] Qatar · Legislative drafter — 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. [marsalqatar.qa](https://marsalqatar.qa/post/80437/%D8%A7%D9%84%D8%A3%D9%85%D8%A7%D9%86%D8%A9-%D8%A7%D9%84%D8%B9%D8%A7%D9%85%D8%A9-%D9%84%D9%85%D8%AC%D9%84%D8%B3-%D8%A7%D9%84%D9%88%D8%B2%D8%B1%D8%A7%D8%A1-%D8%AA), 2026-01-04
- [Deployment] Qatar · Customs officer — 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. [thepeninsulaqatar.com](https://thepeninsulaqatar.com/article/27/01/2026/gac-launches-new-package-of-artificial-intelligence-services), 2026-01-27
- [Deployment] Qatar · Prosecutor (execution of judgments) — 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. [qatar-tribune.com](https://www.qatar-tribune.com/article/247001/latest-news/public-prosecution-launches-ai-powered-self-service-platform-for-inmates), 2026-08-02
- [Deployment] Kuwait · Traffic police officer — 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. [arabtimesonline.com](https://www.arabtimesonline.com/news/ai-to-monitor-public-transport-deliveries-in-kuwait/), 2026-02-02
- [Deployment] Kuwait · Government contact-centre agent — 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. [aljarida.com](https://www.aljarida.com/article/123224), 2026-02-14
- [Deployment] Bahrain · Municipal inspector — 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. [gdnonline.com](https://www.gdnonline.com/Details/1399252/THE-SMART-WAY-AHEAD-AI-powered-cameras-to-guard-Bahrains-public-areas-), 2026-07-01
- [Deployment] United Arab Emirates · Municipal inspector — 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. [gulfnews.com](https://gulfnews.com/uae/government/dubai-launches-ai-cameras-to-catch-litterers-with-fines-of-up-to-dh500-1.500402509), 2026-01-08
- [Deployment] United Arab Emirates · Municipal inspector — 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. [gulfnews.com](https://gulfnews.com/uae/abu-dhabi-expands-smart-inspection-fleet-with-27-ai-powered-vehicles-1.500490036), 2026-03-29
- [Deployment] United Arab Emirates · Contact-centre agent and supervisor — 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. [khaleejtimes.com](https://www.khaleejtimes.com/uae/uae-ministry-finance-ai-call-centre-sentiment-analysis), 2026-05-25
- [Deployment] United Arab Emirates · Weather forecaster — 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. [albayan.ae](https://www.albayan.ae/news/uae/government/1453708), 2026-06-29
- [Regulation] Iraq — 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. [shafaq.com](https://shafaq.com/ar/%D8%A3%D9%85%D9%80%D9%86/%D8%A7%D9%84%D8%AF%D8%A7%D8%AE%D9%84%D9%8A%D8%A9-%D8%A7%D9%84%D8%B9%D8%B1%D8%A7%D9%82%D9%8A%D8%A9-%D8%AA%D8%B7%D9%84%D9%82-%D8%AF%D9%84%D9%8A%D9%84-%D8%B6%D9%88%D8%A7%D8%A8%D8%B7-%D9%88-%D8%AE%D9%84%D8%A7%D9%82%D9%8A%D8%A7%D8%AA-%D8%A7%D8%B3%D8%AA%D8%AE%D8%AF%D8%A7%D9%85-%D8%A7%D9%84%D8%B0%D9%83%D8%A7-%D8%A7%D9%84%D8%A7%D8%B5%D8%B7%D9%86%D8%A7%D8%B9%D9%8A), 2026-03-03
- [Deployment] Iran · Judge — 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. [javanonline.ir](https://www.javanonline.ir/fa/news/1344700/), 2026-02-15
- [Evidence of adoption] Jordan — 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. [asharq.com](https://asharq.com/technology/185388/%D8%B3%D8%B1%D8%A7%D8%AC-%D9%85%D8%B3%D8%A7%D8%B9%D8%AF-%D8%B0%D9%83%D9%8A-%D8%A3%D8%B5%D8%A8%D8%AD-%D8%B1%D9%81%D9%8A%D9%82-%D8%A7%D9%84%D8%AF%D8%B1%D8%A7%D8%B3%D8%A9-%D8%A7%D9%84%D9%8A%D9%88%D9%85%D9%8A-%D9%84%D9%85%D9%84%D9%8A%D9%88%D9%86-%D8%B7%D8%A7%D9%84%D8%A8-%D8%A3%D8%B1%D8%AF%D9%86%D9%8A/), 2026-06-04
- [Regulation] Jordan — 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. [petra.gov.jo](https://petra.gov.jo/adm_program/index.php/ar/news/%D8%A7%D9%84%D8%AD%D9%83%D9%88%D9%85%D8%A9-%D8%AA%D8%B5%D8%AF%D8%B1-%D8%AA%D8%B9%D9%85%D9%8A%D9%85%D8%A7-%D8%A8%D8%B4%D8%A3%D9%86-%D8%A7%D8%B3%D8%AA%D8%AE%D8%AF%D8%A7%D9%85-%D8%A7%D9%84%D8%B0%D9%83%D8%A7%D8%A1-%D8%A7%D9%84%D8%A7%D8%B5%D8%B7%D9%86%D8%A7%D8%B9%D9%8A-%D9%81%D9%8A-%D8%A7%D9%84%D8%AA%D8%B9%D8%A7%D9%85%D9%84-%D9%85%D8%B9-%D8%A7%D9%84%D8%B4%D8%B9%D8%A7%D8%B1%D8%A7%D8%AA-%D9%88%D8%A7%D9%84%D8%B1%D9%85%D9%88%D8%B2-%D8%A7%D9%84%D9%88%D8%B7%D9%86%D9%8A%D8%A9), 2026-07-11
- [Deployment] Egypt · Environmental and building-violations inspector — 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. [dailynewsegypt.com](https://www.dailynewsegypt.com/2026/04/15/egypt-signs-deal-to-deploy-ai-powered-drones-for-environmental-monitoring/), 2026-04-15
- [Regulation] Libya — 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. [eanlibya.com](https://www.eanlibya.com/%D8%A7%D9%84%D8%AF%D8%A8%D9%8A%D8%A8%D8%A9-%D9%8A%D8%B4%D9%87%D8%AF-%D8%A7%D9%86%D8%B7%D9%84%D8%A7%D9%82-%D9%81%D8%B9%D8%A7%D9%84%D9%8A%D8%A7%D8%AA-%D8%A7%D9%84%D9%8A%D9%88%D9%85-%D8%A7%D9%84/), 2026-06-01
- [Deployment] Tunisia · Customs officer — 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. [wearetech.africa](https://www.wearetech.africa/fr/fils/actualites/tech/la-douane-tunisienne-adopte-lintelligence-artificielle-pour-moderniser-ses-controles), 2026-05-19
- [Deployment] Morocco · Tax inspector — 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. [fr.le360.ma](https://fr.le360.ma/economie/controle-fiscal-immobilier-rabat-et-casablanca-pilotes-dun-nouveau-systeme-fonde-sur-lia_OCMJ2K5XF5CY3OYR44PUDRWF3I/), 2026-02-20
- [Deployment] Morocco · Police patrol officer — 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. [fr.le7tv.ma](https://fr.le7tv.ma/2026/07/21/la-dgsn-deploie-les-vehicules-intelligents-amane-et-madar-dans-les-rues-de-rabat/), 2026-07-21
- [Deployment] Mauritania · Civil-service recruitment officer — 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. [saharamedias.net](https://saharamedias.net/254596/), 2026-03-29
- [Deployment] Mauritania · Public enquiry officer — 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. [alakhbar.info](https://www.alakhbar.info/6a3e8331878cf30001d20b87/), 2026-06-26
- [Deployment] Turkey · Tax inspector — 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. [aa.com.tr](https://www.aa.com.tr/tr/ekonomi/naylon-fatura-yapay-zeka-tabanli-kasif-ile-aninda-yakalaniyor/3804685), 2026-01-20
- [Deployment] Turkey · Tax inspector — 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. [aa.com.tr](https://aa.com.tr/tr/ekonomi/mukelleflere-vergi-cezasi-oncesi-erken-uyari-sistemi-geliyor/3959050), 2026-06-07
- [Deployment] Israel · Judge — 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. [globes.co.il](https://www.globes.co.il/news/article.aspx?did=1001542055), 2026-05-06
- [Regulation] Israel — 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. [globes.co.il](https://www.globes.co.il/news/article.aspx?did=1001538140), 2026-03-22
- [Regulation] Israel — 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. [srugim.co.il](https://www.srugim.co.il/1300739-%D7%94%D7%A9%D7%9C%D7%9B%D7%95%D7%AA-ai-%D7%99%D7%A6%D7%97%D7%A7-%D7%A2%D7%9E%D7%99%D7%AA-%D7%9E%D7%9B%D7%AA%D7%91-%D7%A2%D7%95%D7%91%D7%93%D7%99-%D7%9E%D7%A9%D7%A4%D7%98), 2026-03-26
- [Deployment] United States · government customer service agent — 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. [gov.ca.gov](http://www.gov.ca.gov/2026/09/09/government-made-easier-governor-newsom-introduces-askca-a-new-ai-powered-tool-for-californians/), 2026-09-09
- [Evidence of adoption] Canada — 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. [canada.ca](https://www.canada.ca/en/innovation-science-economic-development/news/2026/09/government-of-canada-launches-national-ai-literacy-initiative.html), 2026-09-09
- [Deployment] United States · meeting interpreter — 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. [mv-voice.com](https://www.mv-voice.com/technology/2026/09/09/mountain-view-pilots-ai-translation-services-at-public-meetings/), 2026-09-09
- [Regulation] United States — 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. [fedscoop.com](https://fedscoop.com/national-archives-says-agencies-can-use-some-ai-without-creating-records/), 2026-08-31
- [Deployment] Canada · Court registry officer — 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. [fja.gc.ca](https://www.fja.gc.ca/COVID-19/Repository-of-AI-Initiatives-in-Courts-and-Tribunals-Repertoire-dinitiatives-IA-dans-les-tribunaux-eng.html), 2026
- [Deployment] Canada · Building plans examiner — 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. [canada.constructconnect.com](https://canada.constructconnect.com/dcn/news/technology/2026/07/how-ai-can-answer-the-call-for-faster-permitting), 2026-07
- [Evidence of impact] Canada — 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. [cp24.com](https://www.cp24.com/politics/queens-park/2026/05/12/unsafe-ai-sites-used-by-ontario-government-employees-putting-data-at-risk-auditor/), 2026-05-12
- [Deployment] United States · Building permit reviewer — 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. [smartcitiesdive.com](https://www.smartcitiesdive.com/news/ai-permitting-louisville-kentucky/809355/), 2026-01-12
- [Deployment] United States · Benefits eligibility caseworker — 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. [route-fifty.com](https://www.route-fifty.com/artificial-intelligence/2026/03/michigans-use-ai-process-snap-applications-draws-concerns-about-past-automation-failures/412395/), 2026-03
- [Deployment] United States · Civil servant — 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. [pa.gov](https://www.pa.gov/agencies/oa/newsroom/shapiro-admin-expands-safe-responsible-ai-improve-services), 2026-04-15
- [Deployment] Albania · Police officer — 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. [euronews.al](https://euronews.al/en/rama-and-lamallari-present-smart-city-command-and-control-center-system-detects-threats-in-real-time/), 2026-07-16
- [Regulation] Albania — 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. [rtsh.al](https://rtsh.al/shqiperia-nis-perafrimin-me-be-ne-per-inteligjencen-artificiale-hidhet-per-konsultim-projektligji-i-ri/), 2026-05-29
- [Evidence of adoption] Albania — 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. [tiranadiplomat.com](https://www.tiranadiplomat.com/2026/08/17/inteligjenca-artificiale-ne-sherbim-te-administrates-dhe-qytetareve-prezantohen-kater-sisteme-te-reja/), 2026-08-17
- [Evidence of impact] Albania — 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. [digjitale.com](https://www.digjitale.com/p/signal-71-albania-s-ai-minister-turned-one-and-bought-nothing), 2026-08-18
- [Evidence of impact] Albania — 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. [nbcnews.com](https://www.nbcnews.com/world/europe/albania-artificial-intelligence-government-minister-diella-actor-bisha-rcna258727), 2026-02-12
- [Deployment] Andorra · Lawyer — 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. [laveulliure.ad](https://laveulliure.ad/ca/article/cercador-jurisprudencia-andorra-ia-15-000-sentencies), 2026-01-19
- [Deployment] Andorra · Civil protection officer — 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. [digitalandorra.com](https://digitalandorra.com/societat/la-inquietant-simulacio-de-la-vall-del-madriu-en-flames-per-prevenir-incendis-a-andorra/), 2026-08-05
- [Regulation] Andorra — 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. [alto.ad](https://www.alto.ad/politics/2026/07/andorra-sexual-consent-age-16-penal-code), 2026-07-17
- [Regulation] Benin — 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. [wearetech.africa](https://www.wearetech.africa/fr/fils/actualites/tech/benin-l-usage-non-autorise-de-l-ia-sur-l-image-d-autrui-peut-couter-44-400-d-amende), 2026-06-02
- [Regulation] Benin — 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. [les4verites.bj](https://www.les4verites.bj/un-createur-de-contenus-arrete-apres-une-publication-sur-patrice-talon/), 2026-06-17
- [Deployment] Armenia · Police officer — 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. [armenpress.am](https://armenpress.am/en/article/1258449), 2026-08-20
- [Regulation] Armenia — 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. [lylegal.com](https://lylegal.com/news/development-ai-regulation-armenia-2026), 2026-02-19
- [Deployment] Belarus · Cartographer — 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. [belta.by](https://belta.by/society/view/goskomimuschestvo-iskusstvennyj-intellekt-pomogaet-obnovljat-karty-belarusi-765402-2026/), 2026-02-19
- [Deployment] Belarus · Contact centre operator — 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. [news.by](https://news.by/news/obshchestvo/v-pomosch-kontakt-tsentru-115-bel-prikhodyat-intellektualnye-agenty-boty), 2026-07-23
- [Regulation] Azerbaijan — 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. [caspianlegalcenter.az](https://www.caspianlegalcenter.az/az/news/more/ai-use-crime), 2026-04-21
- [Deployment] Bahamas · Civil servant — 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. [tribune242.com](https://www.tribune242.com/news/2026/jun/16/sebas-pledges-ai-reform-across-the-public-service/), 2026-06-16
- [Regulation] Bosnia and Herzegovina — 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. [sluzbenilist.ba](https://sluzbenilist.ba/page/i/h5YsXqgztz5k76kjn45hmGfk=), 2026-07-10
- [Evidence of adoption] Bosnia and Herzegovina — 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. [bigportal.ba](https://bigportal.ba/chatgpt-usao-u-drzavne-institucije-preko-70-odsto-zaposlenih-koristi-ai/), 2026-07-22
- [Regulation] Burundi — 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. [iwacu-burundi.org](https://www.iwacu-burundi.org/lia-au-coeur-de-la-strategie-numerique-au-burundi/), 2026-04-24
- [Deployment] Barbados · Consumer affairs officer — 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. [barbadostoday.bb](https://barbadostoday.bb/2026/07/15/ask-dealia-app-launches-to-help-find-deals-events/), 2026-07-15
- [Evidence of adoption] Bhutan — 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. [bbs.bt](https://www.bbs.bt/240725/), 2026-04-01
- [Deployment] Croatia · Public administration counter officer — 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. [mpudt.gov.hr](https://mpudt.gov.hr/vijesti-8/uvodjenje-ai-virtualnog-asistenta-na-portal-gov-hr/30900), 2026-06-09
- [Deployment] Croatia · Election commission officer — 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. [bug.hr](https://www.bug.hr/servisi/gradjani-dobili-mogucnost-chata-o-izborima-u-hrvatskoj-uz-alat-glasai-61108), 2026-07-02
- [Deployment] Croatia · Fire monitoring centre dispatcher — 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. [novilist.hr](https://www.novilist.hr/novosti/hrvatska/unaprjeden-sustav-videonadzora-za-rano-otkrivanje-pozara-na-25-novih-lokacija/), 2026-08-25
- [Evidence of adoption] Croatia — 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. [novilist.hr](https://www.novilist.hr/novosti/hrvatska/umjetnu-inteligenciju-u-hrvatskoj-koriste-policija-sudovi-ministarstva-rijetki-imaju-jasno-definirana-pravila), 2026-09-10
- [Deployment] Yemen · Market inspection officer — 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. [sada-alsahel.net](https://www.sada-alsahel.net/news/182308), 2026-07-02
- [Evidence of impact] Yemen — 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. [thenationalnews.com](https://www.thenationalnews.com/news/mena/2026/09/11/how-yemeni-rebels-used-anthropics-ai-software-to-design-and-build-guided-weapons/), 2026-09-11
- [Deployment] Czechia · Tax administration officer — 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. [mf.gov.cz](https://mf.gov.cz/cs/ministerstvo/media/tiskove-zpravy/2026/lidem-s-danemi-nove-pomuze-i-umela-inteligence-fin-63629), 2026-04-09
- [Deployment] Czechia · Lawyer — 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. [ceska-justice.cz](https://www.ceska-justice.cz/2026/02/ustavni-soud-zacal-vyuzivat-umelou-inteligenci/), 2026-02-23
- [Deployment] Czechia · Legal-aid adviser — 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. [ceska-justice.cz](https://www.ceska-justice.cz/2026/05/vrchni-soud-18-tisic-rozhodnuti-chatbot/), 2026-05-13
- [Deployment] Czechia · Municipal clerk — 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. [ratibor.cz](https://ratibor.cz/obec/p2162-ratibor-predstavila-prvniho-ai-starostu-v-cesku), 2026-04-09
- [Deployment] Czechia · Municipal clerk — 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. [zlin.cz](https://zlin.cz/zpravy/na-webu-otrokovic-pomaha-ai-asistent-jeho-ukolem-je-usnadnit-vyhledavani-informaci/), 2026-05-17
- [Regulation] Czechia — 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. [ctu.gov.cz](https://ctu.gov.cz/tiskova-zprava-2.-srpna-2026-jako-klicove-datum-pro-pravidla-transparentnosti-podle-aktu-o-umele), 2026-07-31
- [Deployment] Bermuda · Pensions administrator — 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. [royalgazette.com](https://www.royalgazette.com/local-business/business/article/20260309/government-set-to-build-ai-pension-calculation-chatbot/), 2026-03-09
- [Regulation] Brunei — 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. [thestar.com.my](https://www.thestar.com.my/aseanplus/aseanplus-news/2026/05/01/brunei-ramps-up-efforts-to-ensure-ethical-ai-practices), 2026-05-01
- [Evidence of adoption] Brunei — 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. [bedb.gov.bn](https://www.bedb.gov.bn/post/bedb-launches-national-implementation-of-aim-asean-programme-to-support-msme-growth-through-practica), 2026-01-28
- [Deployment] Bolivia · Judge — 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%. [oxigeno.bo](https://oxigeno.bo/presidente-del-tsj-pide-una-asamblea-constituyente-para-que-se-asigne-mayores-recursos-al-organo-judicial/), 2026-09-07
- [Deployment] Bolivia · Customs broker — 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. [lapatria.bo](https://lapatria.bo/enfoque-nacional/gestion/aduana-nacional-presenta-nuevo-ecosistema-digital/), 2026-08-27
- [Deployment] Bolivia · Prosecutor — 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. [eju.tv](https://eju.tv/2026/07/argo-monitoreo-inteligente-del-ministerio-publico-ya-vigila-a-275-personas-con-detencion-domiciliaria-en-bolivia/), 2026-07-13
- [Deployment] Bolivia · Government communications officer — 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. [montevideo.com.uy](https://www.montevideo.com.uy/Ciencia-y-Tecnologia/Paso-por-paso-video-IA-del-gobierno-boliviano-recrea-la-captura-de-Marset-uc956096), 2026-03-16
- [Evidence of adoption] Bolivia — 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%. [correodelsur.com](https://correodelsur.com/politica/20260826/gobierno-lanza-mas-de-5-000-becas-google-para-capacitar-a-jovenes-bolivianos-en-competencias-digitales.html), 2026-08-26
- [Regulation] Estonia — 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. [valitsus.ee](https://valitsus.ee/uudised/lihtsamates-haldusotsustes-saab-hakata-kasutama-digilahendusi-ja-ai-d), 2026-04-02
- [Evidence of adoption] Estonia — 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. [err.ee](https://www.err.ee/1610087626/riigikantselei-kuulutas-valja-eesti-ai-viie-miljoni-eurose-koolitushanke), 2026-07-28
- [Regulation] Estonia — 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. [riigikohus.ee](https://www.riigikohus.ee/et/uudiste-arhiiv/kohus-trahvis-kaebajaid-tehisaru-hooletu-kasutamise-eest), 2026-06-25
- [Evidence of impact] Estonia — 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. [err.ee](https://www.err.ee/1609992727/avalik-sektor-peab-uha-tihedamini-tegelema-tehisaru-koostatud-poordumistega), 2026-04-10
- [Deployment] Cambodia · Police officer — 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. [akp.gov.kh](https://www.akp.gov.kh/post/detail/373304), 2026-06-17
- [Deployment] Cambodia · Translator — 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. [akp.gov.kh](https://akp.gov.kh/post/detail/358056), 2026-01-02
- [Evidence of adoption] Cambodia — 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. [mptc.gov.kh](https://mptc.gov.kh/en/2026/06/minister-chea-vandeth-presided-over-the-closing-ceremony-of-the-artificial-intelligence-ai-training-program-for-ministry-leadership-and-civil-servants/), 2026-06-11
- [Deployment] Gibraltar · Police officer — 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. [gibraltar.gov.gi](https://www.gibraltar.gov.gi/press-releases/post-treaty-security-posture-keeping-gibraltar-safe-5862026-12232), 2026-07-23
- [Deployment] Gibraltar · Parliamentary researcher — 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. [chronicle.gi](https://www.chronicle.gi/parliament-launches-new-digital-search-tool-for-hansard/), 2026-03-25
- [Deployment] Gibraltar · Debt recovery officer — 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. [chronicle.gi](https://www.chronicle.gi/new-bespoke-ai-system-for-central-arrears-unit-after-recruitment-issues/), 2026-02-02
- [Regulation] Gibraltar — 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. [gra.gi](https://www.gra.gi/uploads/documents/data-protection/Press%20Releases/2026/01%20IRD-pressrelease_230226.pdf), 2026-02-23
- [Regulation] Cayman Islands — 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. [parliament.ky](https://parliament.ky/wp-content/uploads/2026/03/PQ-Answer-83-Artificial-Intelligence-Act-6-March-2026.pdf), 2026-03-06
- [Regulation] Congo — 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. [aci.cg](https://www.aci.cg/congo-recherche-scientifique-une-nouvelle-loi-pour-structurer-le-secteur-et-encadrer-lia/), 2026-04-08
- [Regulation] Guernsey — 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. [odpa.gg](https://www.odpa.gg/news/crown-dependency-data-protection-advisory-generative-ai-image-creation), 2026-02-23
- [Regulation] Vatican City — 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. [press.vatican.va](https://press.vatican.va/content/salastampa/it/bollettino/pubblico/2026/05/16/0416/00814.html), 2026-05-16
- [Deployment] Cook Islands · Parliamentary clerk — 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. [rnz.co.nz](https://www.rnz.co.nz/news/pacific_cook-islands/593589/cook-islands-parliament-launches-ai-agents-to-support-operations), 2026-04-28
- [Deployment] Curaçao · Policy adviser — 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. [ser.cw](https://ser.cw/en/2026/03/26/ser-curacao-embraces-ai-but-draws-a-firm-line-at-human-judgment/), 2026-03-26
- [Regulation] DR Congo — 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. [unesco.org](https://www.unesco.org/fr/articles/la-rdc-se-dote-dune-strategie-nationale-pour-une-intelligence-artificielle-ethique-et-inclusive), 2026-08-12
- [Deployment] Cyprus · Firefighter — 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. [cyprus-mail.com](https://cyprus-mail.com/2026/05/12/cyprus-expands-firefighting-defence-with-aircraft-and-ai-warning-systems), 2026-05-12
- [Deployment] Cyprus · Traffic police officer — 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. [en.sigmalive.com](https://en.sigmalive.com/ai-traffic-cameras-arrive-in-cyprus-what-you-need-to-know/), 2026-07-07
- [Regulation] Cyprus — 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. [oeb.org.cy](https://www.oeb.org.cy/dimosia-diavoyleysi-gia-ton-peri-technitis-noimosynis-nomo-toy-2026/), 2026-08-19
- [Evidence of adoption] Equatorial Guinea — 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. [guineaecuatorialpress.com](https://www.guineaecuatorialpress.com/noticias/el_ministerio_de_transportes_pone_en_marcha_un_ambicioso_programa_de_formacion_en_inteligencia_artificial_y_robotica_junto_a_la_unesco), 2026-07-04
- [Deployment] Hungary · Tax auditor — 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. [nav.gov.hu](https://nav.gov.hu/sajtoszoba/hirek/Vagujhelyi_Ferenc_a_kockazatelemzes_es_mesterseges_intelligencia_tamogatta_kivalasztasra_epulnek_a_2026-os_ellenorzesek), 2026-02-23
- [Deployment] Hungary · Municipal clerk — 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. [pcwplus.hu](https://www.pcwplus.hu/pcwlite/magyarorszag-nagyoroszi-onkormanyzat-aiugyfelszolgalat-chatbot-mesterseges-intelligencia-munipolis-383014.html), 2026-06-02
- [Evidence of impact] Hungary — 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. [qubit.hu](https://qubit.hu/2026/03/04/az-arcfelismero-rendszer-kidob-ot-fotot-kovacs-fotorzs-rabok-az-egyikre-o-lesz-a-gyanusitott), 2026-03-04
- [Deployment] China · Tourist information officer — 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. [globaltimes.cn](https://www.globaltimes.cn/page/202608/1368598.shtml), 2026-08-20
- [Regulation] China — 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. [techxplore.com](https://techxplore.com/news/2026-09-china-court-guidelines-deepfakes-ai.html), 2026-09-07
- [Deployment] Hong Kong · Vehicle licensing officer — 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. [news.gov.hk](https://www.news.gov.hk/eng/2026/08/20260820/20260820_113129_235.html?type=ticker), 2026-08-20
- [Deployment] Pakistan · Pension verification officer — 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. [app.com.pk](https://www.app.com.pk/domestic/nadra-launches-ai-facial-recognition-to-streamline-verification-for-3-mln-pensioners/), 2026-09-05
- [Deployment] Pakistan · Election official — 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. [nation.com.pk](https://www.nation.com.pk/08-Sep-2026/ecp-orders-establishment-ai-wing-sets-deadline-full-digital-office-system), 2026-09-08
- [Deployment] Nepal · Disaster relief coordinator — 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. [kathmandupost.com](https://kathmandupost.com/national/2026/09/06/how-an-it-engineer-used-ai-to-aid-nepal-s-flood-response), 2026-09-06
- [Deployment] Singapore · Small-business adviser — 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. [redhot.sg](https://www.redhot.sg/singapore-businesses-to-get-easier-access-to-government-support-through-ai-powered-bizsg-portal-715038.html), 2026-09-10
- [Deployment] South Korea · Procurement officer — 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. [newswhoplus.com](https://www.newswhoplus.com/news/articleView.html?idxno=69499), 2026-09-11
- [Evidence of adoption] Taiwan — 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. [taipeitimes.com](https://www.taipeitimes.com/News/front/archives/2026/09/01/2003863468), 2026-09-01
- [Deployment] Fiji · tax officer — 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. [fijivillage.com](https://www.fijivillage.com/news/frcs-launches-new-website-marking-a-major-step-towards-digital-transformation-d3a85c/), 2026-04-09
- [Regulation] French Polynesia — 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. [tntvnews.pf](https://www.tntvnews.pf/polynesie/high-tech/ia-dans-ladministration-ce-que-les-agents-du-pays-ont-desormais-le-droit-de-faire-ou-pas/), 2026-02-11
- [Deployment] French Guiana · Tax officer — 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. [franceguyane.fr](https://www.franceguyane.fr/actualite/economie-consommation/impots-2026-lia-pour-simplifier-les-demarches-fiscales-1075849.php), 2026-04-15
- [Deployment] Iceland · civil servant — 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. [stjornarradid.is](https://www.stjornarradid.is/efst-a-baugi/frettir/stok-frett/2026/07/15/Innleiding-gervigreindarlausnar-thvert-a-rikid-/), 2026-07-15
- [Deployment] Georgia · tax accountants and finance officers — 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. [ggi.com](https://www.ggi.com/news/indirect-taxes/the-strategic-advantage-of-georgias-automated-vat-refunds), 2026-04-17
- [Regulation] Gabon — 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. [gabonmediatime.com](https://gabonmediatime.com/gabon-deepfakes-faux-contenus-desinformation-le-gouvernement-encadre-lia-a-marche-forcee/), 2026-04-12
- [Deployment] Liechtenstein · citizen service agent — 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. [presseportal.ch](https://www.presseportal.ch/de/pm/100000148/100940235), 2026-05-26
- [Regulation] Moldova — 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. [logos-pres.md](https://logos-pres.md/en/news/european-commission-has-put-the-brakes-on-the-ai-law-in-moldova/), 2026-02-25
- [Deployment] Gambia · communications professionals, media monitors — 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. [grts.gm](https://www.grts.gm/new-site/public/index.php/news-article-details/news/gambia-launches-west-africas-first-ai-centre-to-fight-fake-news), 2026-04-27
- [Deployment] Hong Kong · Civil servant — 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. [info.gov.hk](https://www.info.gov.hk/gia/general/202608/27/P2026082700593.htm), 2026-08-27
- [Deployment] Kyrgyzstan · member of parliament — 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. [24.kg](https://24.kg/english/369365_Paper_out_E-Kenesh_in_Speaker_of_Parliament_of_Kyrgyzstan_presents_app_for_MPs/), 2026-04-07
- [Regulation] Kyrgyzstan — 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. [ailawradar.com](https://ailawradar.com/jurisdictions/kyrgyzstan), 2026-09-05
- [Deployment] Laos · Civil servant — 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. [kpl.gov.la](https://kpl.gov.la/en/detail.aspx/detail.aspx?id=97734), 2026-04-07
- [Deployment] Macao · civil servant — 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. [macaubusiness.com](https://macaubusiness.com/sar-to-avoid-excessive-regulation-of-artificial-intelligence-secretary-wong/), 2026-04-02
- [Deployment] Macao · food safety inspector — 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. [gcs.gov.mo](https://www.gcs.gov.mo/news/detail/en/N26FEWls18), 2026-06-05
- [Evidence of adoption] Macao — 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. [macaubusiness.com](https://macaubusiness.com/labour-bureau-launches-ai-introductory-courses-for-middle-aged-elderly-residents/), 2026-07-04
- [Regulation] North Macedonia — 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. [new.mia.mk](https://new.mia.mk/al/story/north-macedonia-signs-coe-framework-convention-on-artificial-intelligence), 2026-05-08
- [Deployment] Montenegro · Military command staff — 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. [en.vijesti.me](https://en.vijesti.me/news-b/society/824521/The-Montenegrin-Army-has-developed-its-own-artificial-intelligence-system--the-Digital-Adjutant.), 2026-08-31
- [Deployment] Montenegro · Judges and court staff — 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. [en.vijesti.me](https://en.vijesti.me/news-b/society/801047/Montenegrin-judiciary-launches-first-AI-system-for-case-law), 2026-03-20
- [Deployment] Montenegro · Civil servants and public administration staff — 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. [portalanalitika.me](https://www.portalanalitika.me/clanak/informacije-dostupne-odmah-bez-cekanja-i-administrativnih-prepreka-mju-predstavilo-chatbot-na-portalima-vlade-i-euprave), 2026-04-28
- [Evidence of adoption] Madagascar — 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. [techafricanews.com](https://techafricanews.com/2026/01/29/madagascar-launches-free-15-month-training-program-for-1300-youth-in-digital-sector/), 2026-01-29
- [Regulation] Malaysia — 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. [malaymail.com](https://www.malaymail.com/news/malaysia/2026/09/10/digital-minister-says-malaysias-draft-ai-governance-bill-complete-hopes-to-table-it-by-q1-2027/234656), 2026-09-10
- [Regulation] Maldives — 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. [corporatemaldives.com](https://corporatemaldives.com/govt-to-introduce-ai-regulation-bill-alongside-digital-reform-package/), 2026-04-20
- [Evidence of adoption] Maldives — 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. [mindco.mv](https://www.mindco.mv/newsroom/digital-for-all-campaign-launched), 2026-07-22
- [Regulation] Mali — 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. [wearetech.africa](https://www.wearetech.africa/en/fils-uk/news/tech/six-francophone-west-african-countries-unite-on-ai-governance-framework), 2026-07-09
- [Evidence of impact] Mali — 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. [timbuktu-institute.org](https://timbuktu-institute.org/index.php/toutes-l-actualites/item/1733-in-mali-generative-ai-joins-the-information-war), 2026-09-10
- [Deployment] Serbia · Social worker — 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. [serbiantimes.info](https://serbiantimes.info/en/forget-about-youre-missing-one-document-serbia-launches-its-first-ai-assistant-radoslav-to-help-citizens-access-social-benefits/), 2026-07-23
- [Deployment] Serbia · Administrative clerk — 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. [mreast.eu](https://www.mreast.eu/2026/09/ai-assistant-to-book-id-and-passport-appointments-in-serbia/), 2026-09-08
- [Deployment] Serbia · Police officer — 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. [balkaninsight.com](https://balkaninsight.com/2026/06/01/serbia-grows-its-facial-recognition-arsenal-amid-legality-concerns/bi/), 2026-06-01
- [Regulation] Slovakia — 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. [teraz.sk](https://www.teraz.sk/spravy/vlada-dohlad-nad-ai-ma-koordinovat/987606-clanok.html), 2026-08-26
- [Regulation] Slovenia — 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. [gov.si](https://www.gov.si/novice/2026-08-03-zacela-so-se-uporabljati-pravila-o-preglednosti-sistemov-umetne-inteligence/), 2026-08-03
- [Deployment] Albania · School security officer — 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. [inalbania.info](https://www.inalbania.info/pm-rama-20-schools-in-tirana-equipped-with-smart-cameras-to-strengthen-security/), 31 Aug 2026
- [Deployment] Belarus · Customs officer — 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. [belta.by](https://belta.by/society/view/ot-elektronnoj-tamozhni-k-intellektualnoj-v-gtk-rasskazali-o-novom-etape-razvitija-800309-2026/), 10 Sep 2026
- [Deployment] Belgium · Police officer — 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. [vrt.be](https://www.vrt.be/vrtnws/nl/2026/09/03/gent-politie-chatbot-24-uur-digitale-inspecteur-artificiele-inte/), 3 Sep 2026
- [Regulation] Czechia — 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. [lupa.cz](https://www.lupa.cz/aktuality/ctu-rozjizdi-dohled-nad-umelou-inteligenci-nove-oddeleni-zahajilo-provoz/), 1 Sep 2026
- [Regulation] Hungary — 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. [gdpr.blog.hu](https://gdpr.blog.hu/2026/09/02/new_set-up_in_the_supervision_of_ai_systems_in_hungary), 2 Sep 2026
- [Deployment] Italy · Municipal officer — 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. [genovaquotidiana.com](https://genovaquotidiana.com/2026/09/01/pieve-ligure-porta-43-intelligenze-artificiali-negli-uffici-controlleranno-atti-e-norme/), 1 Sep 2026
- [Deployment] Réunion · tax inspector — 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. [lequotidien.re](https://www.lequotidien.re/article/societe/2026/04/15/Une-lutte-contre-la-fraude-renforcee-557157), 2026-04-15
- [Deployment] Switzerland · public administration officer — 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. [it-markt.ch](https://www.it-markt.ch/news/2026-03-23/kanton-tessin-setzt-fuer-uebersetzungen-auf-apertus), 2026-03-23
- [Deployment] Switzerland · meteorologist — 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. [nau.ch](https://www.nau.ch/news/forschung/meteoschweiz-setzt-auf-ki-fur-prazisere-gewittervorhersagen-67167141), 2026-08-27
- [Deployment] New Caledonia · employment officer — 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. [la1ere.franceinfo.fr](https://la1ere.franceinfo.fr/nouvellecaledonie/le-gouvernement-presente-une-nouvelle-plateforme-destinee-a-relancer-le-marche-de-l-emploi-1680903.html), 2026-03-14
- [Deployment] Ukraine · citizen — 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. [biometricupdate.com](https://www.biometricupdate.com/202607/ukraine-expands-diia-ai-with-voice-based-access-to-government-services), 2026-07-21
- [Deployment] Ukraine · policy analyst — 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. [kyivpost.com](https://www.kyivpost.com/post/74338), 2026-04-20
- [Deployment] United Kingdom · civil servant — 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. [gds.blog.gov.uk](https://gds.blog.gov.uk/2026/05/14/gov-uk-chat-launches/), 2026-05-14
- [Deployment] United Kingdom · planning officer — 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. [bbc.co.uk](https://www.bbc.co.uk/news/articles/c20yx56n133o), 2026-06-17
- [Deployment] United Kingdom · police officer — 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. [theguardian.com](https://www.theguardian.com/technology/2026/jun/23/met-to-expand-live-facial-recognition-central-london), 2026-06-23
- [Deployment] Tajikistan · Tax officers — 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. [asiaplus.news](https://asiaplus.news/en/2026/03/10/ai-in-tajikistans-tax-system-how-technology-is-changing-interactions-with-citizens-and-businesses/), 2026-03-10
- [Deployment] Tajikistan · Civil servants — 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. [en.avesta.tj](https://en.avesta.tj/2026/01/09/tajikistan-launches-national-ai-gov-initiative-to-modernize-public-administration/), 2026-01-09
- [Deployment] Tajikistan · Pension administrators — 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. [asiaplus.news](https://asiaplus.news/en/2026/07/30/face-id-for-pensioners-to-become-permanent-tajikistan-completes-large-scale-re-registration/), 2026-07-30
- [Regulation] Papua New Guinea — 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. [abc.net.au](https://www.abc.net.au/news/2026-07-21/new-laws-in-png-target-ai-deepfakes/106928414), 2026-07-21
- [Deployment] Zimbabwe · cybersecurity analyst — 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. [zimetro.co.zw](https://zimetro.co.zw/us30m-lost-to-mobile-fraud-as-zimbabwe-fights-back-with-ai/), 2026-04-08
- [Deployment] Venezuela · disaster response official — 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. [mincyt.gob.ve](https://mincyt.gob.ve/ia-permite-identificar-afectaciones-por-sismos-del-24-de-junio-mediante-analisis-satelital-y-drones/), 2026-07-25
- [Regulation] Venezuela — 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. [mincyt.gob.ve](https://mincyt.gob.ve/venezuela-presenta-el-codigo-de-etica-para-el-desarrollo-y-aplicacion-responsable-de-la-ia/), 2026-02-19
- [Deployment] Somalia · humanitarian aid worker — 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. [news.un.org](https://news.un.org/en/story/2026/08/1168066), 2026-08-03
- [Regulation] Åland Islands — 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. [lagtinget.ax](https://www.lagtinget.ax/arenden/60082/60525), 2026-06-03
- [Regulation] Falkland Islands (Malvinas) — 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. [fitv.co.fk](https://fitv.co.fk/news-and-events/new-laws-criminalising-deepfake-intimate-images-in-the-falklands/), 2026-06-18
- [Deployment] French Southern Territories · conservation manager — 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. [spaceclimateobservatory.org](https://www.spaceclimateobservatory.org/cartovege2-mapping-unique-biodiversity-sub-antarctic-islands), 2026-06-17
- [Evidence of adoption] Guam — 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. [pacificislandtimes.com](https://www.pacificislandtimes.com/post/guam-s-labor-department-begins-ai-training), 2026-02-02
- [Evidence of adoption] Jamaica — 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. [opm.gov.jm](https://opm.gov.jm/government-launches-growing-ai-innovation-and-national-skills-programme/), 2026-06-03
- [Deployment] Malta · Court reporter — 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. [lovinmalta.com](https://lovinmalta.com/lifestyle/technology/court-and-parliament-sittings-could-soon-be-transcribed-by-ai-in-maltese/), 2026-05-14
- [Regulation] Malta — 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. [govinsider.asia](https://govinsider.asia/intl-en/article/malta-forms-ai-governance-council-to-put-people-at-the-centre-of-ai-adoption), 2026-07-30
- [Deployment] Mayotte · border security officer — 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. [radiofrance.fr](https://www.radiofrance.fr/franceinter/podcasts/l-info-de-france-inter/l-info-de-france-inter-2541075), 2026-08-26
- [Deployment] Sint Maarten (Dutch part) · public auditor — 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. [smn-news.com](https://smn-news.com/index.php/st-maarten-st-martin-news/49837-audit-intelligence-launched-to-support-informed-public-oversight.html), 2026-01-26
- [Deployment] Sint Maarten (Dutch part) · civil servant — 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. [721news.com](https://www.721news.com/2026/09/governors-address-2026-delivered-on-the-occasion-of-the-opening-of-parliamentary-year-2026-2027/), 2026-09-09
- [Deployment] Turkmenistan · Customs declarant — 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. [customs.gov.tm](https://www.customs.gov.tm/en/news/1086), 2026-06-01
- [Evidence of adoption] Malta — 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. [euronews.com](https://www.euronews.com/next/2026/05/16/malta-offers-free-chatgpt-plus-access-to-its-citizens-through-a-national-ai-program), 2026-05-16
- [Deployment] Anguilla · Parliamentary staff — 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. [theanguillian.com](https://theanguillian.com/2025/12/house-of-assembly-launches-its-hansard-ai-project-and-new-website/), 2026-01-13
- [Regulation] Algeria — 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. [ailawradar.com](https://ailawradar.com/jurisdictions/algeria), 2025-07-24
- [Deployment] Canada · civil servant — 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. [globalnews.ca](https://globalnews.ca/news/11525236/ontario-government-artificial-intelligence-strategy/), 2025-11-14
- [Deployment] Canada · customer service agent — 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. [ville.mont-royal.qc.ca](https://www.ville.mont-royal.qc.ca/en/news/miscellaneous/x-3), 2025-06-25
- [Regulation] Canada — 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. [cbc.ca](https://www.cbc.ca/news/politics/federal-government-public-ai-register-9.6995056), 2025-11-28
- [Evidence of adoption] Canada — 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. [newswire.ca](https://www.newswire.ca/news-releases/half-of-public-servants-who-use-ai-rely-on-publicly-available-ai-tools-in-their-job-exposing-governments-to-a-wide-range-of-risks-824090878.html), 2025-09-18
- [Deployment] Canada · public servant — 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. [cbc.ca](https://www.cbc.ca/news/canada/ottawa/federal-government-to-pilot-ai-translation-tool-in-6-departments-1.7647183), 2025-09-30
- [Deployment] Canada · citizen — 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. [cbc.ca](https://www.cbc.ca/news/canada/windsor/ai-artificial-intelligence-chat-bot-windsor-9.7023033), 2025-12-19
- [Regulation] Canada — 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. [quebec.ca](https://www.quebec.ca/nouvelles/actualites/details/utilisation-de-lintelligence-artificielle-dans-ladministration-publique-le-gouvernement-definit-le-cadre-et-les-principes-devant-etre-respectes-par-les-organismes-publics-67850), 2025-12-19
- [Deployment] Canada · Police officer — 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. [nationalpost.com](https://nationalpost.com/news/politics/rcmp-quietly-testing-ai-drafted-reports-from-body-camera-audio), 2026-04-06
- [Regulation] Canada — 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. [www-u.fct-cf.gc.ca](https://www-u.fct-cf.gc.ca/en/pages/law-and-practice/artificial-intelligence), 2025-09-29
- [Deployment] Ethiopia · customs officer — 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. [techreviewafrica.com](https://techreviewafrica.com/news/2883/ethiopian-institute-of-artificial-intelligence-unveils-new-digital-customs-systems), 2025-10-10
- [Deployment] Ethiopia · police officer — 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. [addisfortune.news](https://addisfortune.news/ai-institute-federal-police-ink-deal-for-smart-station), 2025-07-13
- [Deployment] Ethiopia · judge — 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. [techreviewafrica.com](https://techreviewafrica.com/news/3262/ethiopian-ai-institute-and-supreme-court-of-oromia-partner-to-develop-afaan-oromo-court-transcription-system), 2025-11-26

## Healthcare & Life Sciences

- [Deployment] United Kingdom · Radiologist — 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. [resultsense.com](https://resultsense.com/news/2026-08-26-qvh-chest-xray-ai/), 26 Aug 2026
- [Deployment] United States · Radiologist — 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. [itnonline.com](https://itnonline.com/content/osf-healthcare-rapidai-expand-enterprise-ai-partnership), 10 Aug 2026
- [Deployment] United States · Radiologist — 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. [news.uhhospitals.org](https://news.uhhospitals.org/news-releases/articles/2026/08/twelve-us-health-systems-and-aidoc-unite), 11 Aug 2026
- [Deployment] Denmark · Drug researcher — 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. [globenewswire.com](https://www.globenewswire.com/news-release/2026/08/10/3341794/0/en/novo-nordisk-and-aws-enter-strategic-partnership-to-accelerate-drug-discovery-through-ai.html), 10 Aug 2026
- [Deployment] United States · Clinician — 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. [prnewswire.com](https://www.prnewswire.com/news-releases/lcmc-health-announces-strategic-ai-partnership-with-qualified-health-to-enhance-patient-care-and-safety-302848687.html), 11 Aug 2026
- [Deployment] United Arab Emirates · Radiologist — 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. [healthcaremea.com](https://www.healthcaremea.com/2026/08/26/uae-hospitals-expand-ai-and-connected-technologies-to-transform-patient-care/), 26 Aug 2026
- [Deployment] United Arab Emirates · Clinician — 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. [healthcaremea.com](https://www.healthcaremea.com/2026/08/26/uae-hospitals-expand-ai-and-connected-technologies-to-transform-patient-care/), 26 Aug 2026
- [Deployment] United Arab Emirates · Clinician — 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. [healthcaremea.com](https://www.healthcaremea.com/2026/08/26/uae-hospitals-expand-ai-and-connected-technologies-to-transform-patient-care/), 26 Aug 2026
- [Deployment] United States · Customer-service agent — 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. [hitconsultant.net](https://hitconsultant.net/2026/07/29/musc-health-expands-soundhound-ai-agent-emily-pharmacy-epic/), Aug 2026
- [Deployment] Italy · Stroke clinician — 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. [varesenoi.it](https://www.varesenoi.it/2026/08/24/leggi-notizia/argomenti/salute-3/articolo/allospedale-di-circolo-lintelligenza-artificiale-accelera-la-diagnosi-e-il-trattamento-dellictus.html), 2026-08-24
- [Deployment] United Kingdom · Clinician — 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. [resultsense.com](https://www.resultsense.com/news/2026-08-27-thinksono-dvt-chelwest/), 2026-08-27
- [Deployment] United Kingdom · Prescriber — 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. [htn.co.uk](https://htn.co.uk/2026/08/18/calderdale-and-huddersfield-to-launch-epr-tool-for-clinical-decision-support-for-medicines-management/), 2026-08-18
- [Deployment] United Kingdom · Clinician — 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. [htn.co.uk](https://htn.co.uk/2026/08/05/kingston-and-richmond-ambient-voice-rollout/), 2026-08-05
- [Deployment] New Zealand · ED clinician — 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. [globalgovernmentforum.com](https://www.globalgovernmentforum.com/ai-innovations-in-health-new-zealand-to-roll-out-ai-medical-scribe-and-saudi-arabia-launches-digital-health-coach/), 2026-08-15
- [Deployment] Singapore · Endoscopist — 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%. [taiwannews.com.tw](https://www.taiwannews.com.tw/news/6414125), 2026-08-04
- [Deployment] Canada · Clinician — 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. [unityhealth.to](https://unityhealth.to/2026/08/gemini-launches-ai-assisted-trial-prevent-delirium-across-13-ontario-hospitals/), 2026-08-04
- [Evidence of adoption] United States — Reported AI use among U.S. physicians reached 81% in early 2026, up from 38% in 2023. It more than doubled in three years. [ama-assn.org](https://www.ama-assn.org/practice-management/digital-health/more-80-physicians-use-ai-professionally-ama-survey), 2026
- [Evidence of adoption] United States — 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. [fda.gov](https://www.fda.gov/medical-devices/software-medical-device-samd/artificial-intelligence-enabled-medical-devices), 2026
- [Evidence of adoption] United States — 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. [rockhealth.com](https://rockhealth.com/insights/2025-year-end-digital-health-funding-overview-a-tale-of-two-markets/), 2025
- [Deployment] United States · Clinician — 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. [catalyst.nejm.org](https://catalyst.nejm.org/doi/abs/10.1056/CAT.25.0040), 2026
- [Evidence of impact] Sweden — 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%. [thelancet.com](https://www.thelancet.com/journals/lancet/article/PIIS0140-6736(25)02464-X/abstract), 2026
- [Evidence of impact] Poland — 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. [statnews.com](https://www.statnews.com/2025/08/12/ai-deskilling-doctors-colonoscopy-study-lancet/), 2025
- [Deployment] United Kingdom · Stroke clinician — 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. [med-techinsights.com](https://med-techinsights.com/2025/09/03/stroke-ai-software-hailed-as-revolutionary-after-new-study/), 2025
- [Deployment] China · Drug researcher — Insilico Medicine's Rentosertib, the first AI-designed drug (target and molecule from generative AI) to report a Phase 2a trial, in idiopathic pulmonary fibrosis (Nature Medicine, June 2025), entered Phase III in July 2026. The randomised, double-blind, placebo-controlled pivotal trial enrols 320 patients over 52 weeks (NCT07687459), a registrational first for a fully AI-originated drug. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC12353801/), 2025
- [Deployment] United States · Clinician — 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. [fortune.com](https://fortune.com/2024/10/26/openai-transcription-tool-whisper-hallucination-rate-ai-tools-hospitals-patients-doctors), 2024
- [Evidence of impact] Global — 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. [nature.com](https://www.nature.com/articles/s41591-026-04503-6), 2026
- [Deployment] United States · Clinician / patient — 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. [statnews.com](https://www.statnews.com/2026/07/02/fda-clearance-raises-questions-updoc-use-generative-ai-diabetes-treatment/), 2026
- [Evidence of impact] United States — 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. [healthjournalism.org](https://healthjournalism.org/blog/2026/03/ai-diagnostic-risks-top-ecris-2026-patient-safety-concerns/), 2026
- [Evidence of impact] Global — 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. [nature.com](https://www.nature.com/articles/s41598-026-39051-z), 2026
- [Deployment] Brazil · intensive care clinician — 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. [gov.br](https://www.gov.br/saude/pt-br/assuntos/noticias-ms/2026/agosto/ministro-da-saude-participa-de-simposio-sobre-hospitais-inteligentes-e-uso-da-inteligencia-artificial-no-cuidado), 2026-08-05
- [Deployment] South Korea · Claims reviewer — 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. [koreaittimes.com](https://www.koreaittimes.com/news/articleView.html?idxno=156820), 2026-09-01
- [Deployment] France · Doctor — 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. [alliancy.fr](https://alliancy.fr/lhopital-foch-deploie-un-assistant-medical-dia-pour-automatiser-les-comptes-rendus-de-consultation-0d0791e2-3905-4bed-b710-ce46cfde8969), 2026-01-29
- [Deployment] France · Doctor — 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. [caducee.net](https://www.caducee.net/actualite-medicale/16924/medgpt-au-chu-de-poitiers-l-ia-medicale-entre-promesse-clinique-et-surveillance.html), 2026-06-12
- [Deployment] Mexico · Radiologist — 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. [razon.com.mx](https://www.razon.com.mx/mexico/2026/03/09/hospital-oncologico-para-la-mujer-de-la-cdmx-del-imss-bienestar-atendera-a-500-mil-mujeres-al-ano-lo-inaugura-csp/), 2026-03-09
- [Deployment] Chile · Surgical theatre coordinator — 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. [g5noticias.cl](https://g5noticias.cl/2026/07/29/hospital-biprovincial-optimiza-capacidad-quirurgica-a-pacientes-a-traves-de-nuevo-sistema-informatico/), 2026-07-20
- [Deployment] Morocco · Midwife — 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. [aujourdhui.ma](https://aujourdhui.ma/societe/diagnostic-precoce-lintelligence-artificielle-sinvite-dans-les-hopitaux-publics), 2026-06-01
- [Evidence of adoption] Canada — 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. [cbc.ca](https://www.cbc.ca/news/canada/nova-scotia/ai-scribe-healthcare-9.7053460), 2026-01
- [Deployment] Kenya · Health insurance claims officer — 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. [capitalfm.africa](https://capitalfm.africa/ai-detects-sha-fraud-after-member-registered-with-381-children/), 2026-01-29
- [Deployment] Uganda · Midwife — 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. [newvision.co.ug](https://www.newvision.co.ug/category/health/uganda-deploys-ai-powered-portable-ultrasound-NV_232896_052026), 2026-04-29
- [Deployment] Malaysia · Elderly care worker — 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. [mediaselangor.com](https://mediaselangor.com/en/2026/01/340028), 2026-01-19
- [Deployment] Kazakhstan · Radiologist — 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. [kazlenta.kz](https://kazlenta.kz/119685-kak-iskusstvennyy-intellekt-pomogaet-vyyavlyat-insult-v-shymkente.html), 2026-08-31
- [Deployment] Australia · Hospital clinician — 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. [pulseit.news](https://www.pulseit.news/australian-digital-health/metro-south-rolls-out-heidi-ai-scribe/), 2026-09-02
- [Deployment] Vietnam · Radiologist — 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. [bachmai.gov.vn](https://bachmai.gov.vn/bai-viet/tam-soat-ung-thu-phoi-mien-phi-tai-benh-vien-bach-mai-phat-hien-2-truong-hop-nghi-ngo-ac-tinh-ngay-trong-ngay-dau-tien?id=061c2c94-ee64-495e-851c-1b2ad36e2d50), 2026-05-13
- [Deployment] South Korea · Hospital doctor — 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. [newsis.com](https://www.newsis.com/view/NISX20260528_0003648191), 2026-05-29
- [Deployment] Thailand · Radiologist — 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. [engineeringtoday.net](https://www.engineeringtoday.net/thai-ai-xray-siriraj-nhso/), 2026-08-14
- [Deployment] Sri Lanka · Radiologist — 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. [newswire.lk](https://www.newswire.lk/2026/03/09/ai-enabled-mri-machine-installed-at-national-hospital-colombo/), 2026-03-09
- [Deployment] Indonesia · Health insurance claims verifier — 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. [news.detik.com](https://news.detik.com/berita/d-8338381/optimalkan-layanan-jkn-bpjs-kesehatan-hadirkan-4-inovasi-berbasis-ai), 2026-02-03
- [Deployment] Belgium · Radiologist — 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. [vrt.be](https://www.vrt.be/vrtnws/nl/2026/02/27/gent-ai-uz-universiteit-universitair-ziekenhuis-manager-toepassi/), 2026-02-27
- [Evidence of adoption] Czechia — 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. [euro.cz](https://www.euro.cz/clanky/umela-inteligence-se-zabydluje-v-ceskych-nemocnicich-doktori-ji-pouzivaji-hlavne-v-radiologii-a-kardiologii/), 2026-04
- [Deployment] Israel · Family physician — 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. [healthy.walla.co.il](https://healthy.walla.co.il/item/3836051), 2026-05-06
- [Evidence of impact] Oman — 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. [omannews.gov.om](https://www.omannews.gov.om/topics/ar/3/show/472501), 2026-08-24
- [Deployment] Romania · Hospital receptionist — 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. [mesagerulneamt.ro](https://mesagerulneamt.ro/2026/09/fobia-purtatorilor-de-cuvant-virtuali-bantuie-si-sju-piatra-neamt-holograma-la-policlinica-pacienti-la-ghiseu/), 2026-09-03
- [Deployment] Russia · General practitioner — 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. [m24.ru](https://www.m24.ru/news/09092026/939280), 2026-09-09
- [Deployment] Spain · Radiologist — 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. [gacetinmadrid.com](https://gacetinmadrid.com/2026/05/04/madrid-implantara-artificial-hospitales-diagnostico-canceres-prostata-mama/), 2026-05-04
- [Deployment] Benin · Medical laboratory technician — 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. [koreatimes.co.kr](https://www.koreatimes.co.kr/business/companies/20260907/benin-enlists-korean-developed-ai-in-nationwide-battle-against-malaria), 2026-09-07
- [Deployment] Brazil · Physician — 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. [jornalgrandebahia.com.br](https://jornalgrandebahia.com.br/2026/09/governo-ronaldo-adota-inteligencia-artificial-para-apoiar-decisoes-clinicas-na-rede-municipal-de-saude-de-feira-de-santana/), 2026-09-03
- [Deployment] Brazil · Intensive-care clinician — 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. [agenciagov.ebc.com.br](https://agenciagov.ebc.com.br/noticias/202609/ministerio-da-saude-apresenta-central-de-comando-das-utis-inteligentes-do-sus), 2026-09-03
- [Deployment] Costa Rica · Radiologist — 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. [infobae.com](https://www.infobae.com/costa-rica/2026/09/07/hospital-de-costa-rica-suma-ultrasonidos-con-inteligencia-artificial-para-mamas-y-tiroides/), 2026-09-04
- [Deployment] Finland · Dental clinic appointment staff — 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. [pohde.fi](https://pohde.fi/ajankohtaista/uutiset/suun-terveydenhuollon-ajan-voi-nyt-perua-tai-siirtaa-tekoalyapurin-avulla-kahdeksalla-uudella-paikkakunnalla/), 2026-05-27
- [Deployment] Denmark · Emergency nurse and junior doctor — 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. [via.ritzau.dk](https://via.ritzau.dk/pressemeddelelse/14903736/er-den-braekket-patienter-kan-komme-hurtigere-hjem-fra-akutmodtagelsen-med-kunstig-intelligens?publisherId=13562843&lang=da), 2026-04-29
- [Deployment] Lithuania · Pathologist — 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. [15min.lt](https://www.15min.lt/sveikata/naujiena/sveikatos-naujienos/dirbtinis-intelektas-kovoje-su-kruties-veziu-lietuvoje-pradedama-taikyti-itin-tiksli-diagnostika-1696-2677086), 2026-05-06
- [Deployment] Bulgaria · Urologist — 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. [news.bg](https://news.bg/health/vma-izpolzva-inovativna-ai-navigatsiya-v-urologichnite-interventsii.html), 2026-07-01
- [Deployment] Norway · Hospital clinician — 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. [helse-sorost.no](https://www.helse-sorost.no/497a50/siteassets/documents/styret/styremoter/2026/0423/064-2026-driftsorienteringer-fra-administrerende-direktor.pdf), 2026-04-16
- [Evidence of adoption] Netherlands — 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. [mxi.nl](https://mxi.nl/kennis/752/ai-monitor-ziekenhuizen-2026), 2026-03-16
- [Regulation] Ireland — 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. [rte.ie](https://www.rte.ie/news/2026/0729/1585587-health-ai-use/), 2026-07-29
- [Deployment] Kenya · Community health promoter — 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. [techweez.com](https://techweez.com/2026/03/05/kenya-ai-community-health-partnership-casual-foundry/), 2026-03-05
- [Deployment] Tanzania · Malaria vector-control officer — 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. [thecitizen.co.tz](https://www.thecitizen.co.tz/tanzania/zanzibar/how-drones-and-ai-are-powering-zanzibar-s-malaria-war-5458282), 2026-05-13
- [Deployment] Ethiopia · Health extension worker — 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. [thereporterethiopia.com](https://www.thereporterethiopia.com/50178/), 2026-04-11
- [Deployment] Tanzania · Medical laboratory scientist — 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. [thecitizen.co.tz](https://www.thecitizen.co.tz/tanzania/news/national/jakaya-kikwete-cardiac-institute-launches-ai-and-robotic-laboratory-5418006), 2026-04-09
- [Deployment] Uganda · Cervical screening health worker — 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. [nilepost.co.ug](https://nilepost.co.ug/health/323392/govt-strengthens-cervical-cancer-prevention-with-ai-screening-and-shs11bn-korean-support), 2026-02-26
- [Deployment] Cote d'Ivoire · District clinician treating childhood TB — 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. [aip.ci](https://www.aip.ci/cote-divoire-aip-un-outil-numerique-innovant-experimente-a-sassandra-pour-la-lutte-contre-la-tuberculose/), 2026-06-18
- [Deployment] Costa Rica · Waiting-list administrator — 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. [delfino.cr](https://delfino.cr/2026/05/ccss-usara-inteligencia-artificial-para-depurar-listas-de-espera), 2026-05-08
- [Deployment] Panama · Epidemiological surveillance officer — 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. [minsa.gob.pa](https://www.minsa.gob.pa/noticia/minsa-implementa-plan-piloto-vanessa-para-prevenir-el-dengue-en-san-miguelito), 2026-01-08
- [Deployment] El Salvador · Physician — 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. [lanoticiasv.com](https://lanoticiasv.com/bukele-lanza-etapa-2-de-doctor-sv-con-ia-para-atender-enfermedades-cronicas/), 2026-04-15
- [Deployment] Puerto Rico · Radiologist — 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. [metro.pr](https://www.metro.pr/estilo-vida/2026/07/20/centro-medico-estrena-ct-scan-con-inteligencia-artificial/?outputType=amp), 2026-07-20
- [Deployment] Colombia · Public-health suicide-prevention officer — 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. [infobae.com](https://www.infobae.com/colombia/2026/04/01/bogota-utilizara-inteligencia-artificial-para-detectar-riesgos-para-la-salud-mental-en-redes-sociales-asi-funcionara-la-nueva-estrategia/), 2026-04-01
- [Deployment] Argentina · Emergency triage nurse — 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. [rionegro.com.ar](https://www.rionegro.com.ar/politica/como-funcionara-el-sistema-de-inteligencia-artificial-que-rio-negro-probara-en-las-guardias-hospitalarias/), 2026-03-15
- [Deployment] Peru · Radiologist — 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. [andina.pe](https://andina.pe/agencia/noticia-minsa-moderniza-radiologia-hospitales-lima-equipos-tecnologicos-basados-ia-1059047.aspx), 2026-01-12
- [Deployment] Brazil · Oncologist — 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. [exame.com](https://exame.com/inteligencia-artificial/ia-do-google-chega-ao-sus-do-parana-para-ajudar-medicos-a-identificar-tratamentos-contra-o-cancer/), 2026-07-08
- [Deployment] Brazil · Regulating physician — 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. [prefeitura.poa.br](https://prefeitura.poa.br/procempa/noticias/inteligencia-artificial-qualifica-gestao-de-internacoes-do-sus-na-prefeitura-de), 2026-07-28
- [Deployment] Mexico · Non-radiologist physician — 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. [gob.mx](https://www.gob.mx/salud/prensa/006-iner-utiliza-inteligencia-artificial-para-impulsar-deteccion-temprana-de-cancer-de-pulmon), 2026-01-15
- [Deployment] Mexico · Breast radiologist — 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. [infobae.com](https://www.infobae.com/mexico/2026/06/02/issste-interpretara-mastografias-con-ayuda-de-la-ia-y-a-distancia-para-agilizar-deteccion-de-cancer-de-mama/), 2026-06-02
- [Deployment] United Arab Emirates · Embryologist — 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. [gulfnews.com](https://gulfnews.com/uae/health/sehas-corniche-fertility-centre-deploys-ai-embryo-selection-to-boost-ivf-success-across-purehealth-network-1.500552671), 2026-05-25
- [Deployment] Tunisia · Primary-care doctor — 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. [leconomistemaghrebin.com](https://www.leconomistemaghrebin.com/2026/07/21/sante-numerique-la-tunisie-mise-sur-lintelligence-artificielle-pour-transformer-son-systeme-de-soins/), 2026-07-21
- [Evidence of adoption] Turkey — 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. [aa.com.tr](https://www.aa.com.tr/tr/saglik/saglik-bakanligi-yapay-zeka-destekli-mamografi-tarama-raporlama-sistemi-gelistirdi/3891407), 2026-04-04
- [Deployment] Israel · Emergency physician — 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. [hitconsultant.net](https://hitconsultant.net/2026/02/26/sheba-medical-center-arc-smarter-ambient-ai-emergency-department/), 2026-02-26
- [Deployment] Canada · pharmacist — 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. [loblaw.ca](https://www.loblaw.ca/en/health-questions-ask-away-pc-health-launches-new-built-in-canada-for-canadians-ai-powered-health-chat/), 2026-09-10
- [Deployment] United States · clinician — 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. [hcinnovationgroup.com](https://www.hcinnovationgroup.com/clinical-it/electronic-health-record-electronic-medical-record-ehr-emr/news/55403391/with-epic-go-live-adventist-health-taking-advantage-of-ai-tools), 2026-09-08
- [Evidence of impact] Canada — 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. [globalnews.ca](https://globalnews.ca/news/11844349/ontario-auditor-general-ai-usage/), 2026-05-12
- [Evidence of impact] Canada — 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. [newswire.ca](https://www.newswire.ca/news-releases/canada-health-infoway-releases-results-of-national-ai-scribe-program-demonstrating-meaningful-benefits-for-primary-care-clinicians-862493354.html), 2026-06-23
- [Evidence of impact] Lebanon — 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. [pubmed.ncbi.nlm.nih.gov](https://pubmed.ncbi.nlm.nih.gov/42415044/), 2026-07-07
- [Evidence of adoption] Afghanistan — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC13461032/), 2026-07-09
- [Deployment] Antigua and Barbuda · Radiographer — 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. [antigua.news](https://antigua.news/2026/01/19/new-diagnostic-centre-offers-early-mri-access-amid-healthcare-backlog/), 2026-01-19
- [Deployment] Benin · Vector control officer — 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. [gouv.bj](https://www.gouv.bj/article/3570/lutte-contre-paludisme-demarrage-pulverisation-drone-dans-communes-benin/), 2026-07-01
- [Evidence of adoption] Sudan — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC12937154/), 2026-02-15
- [Deployment] Belarus · Radiologist — 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. [belnovosti.by](https://www.belnovosti.by/minsk/stolichnye-rentgenologi-budut-ispolzovat-ii-kommentariy-minzdrava), 2026-03-23
- [Deployment] Aruba · Radiologist — 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. [solodipueblo.com](https://solodipueblo.com/implementacion-di-metodo-nobo-pun-miho-cuido-di-pashent-cuido-reforsa-cu-avance-den-diagnosis-di-prostaat-na-hoh/), 2026-06-04
- [Evidence of impact] Syria — 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. [nature.com](https://www.nature.com/articles/s41598-026-57032-0), 2026-06-06
- [Evidence of impact] Syria — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC13380955/), 2026-06-19
- [Evidence of impact] Syria — 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. [researchsquare.com](https://www.researchsquare.com/article/rs-10297530/v1), 2026-08-10
- [Evidence of adoption] Syria — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC12979924/), 2026-03-10
- [Deployment] Azerbaijan · Radiologist — 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. [far.az](https://far.az/114196-azerbaycanda-xesteliklerin-ashkarlanmasinda-suni-intellektden-istifade-edilecek.html), 2026-09-08
- [Deployment] Bhutan · Clinician — 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. [businesswire.com](https://www.businesswire.com/news/home/20260608871442/en/Cardiolife-Partners-with-Bhutan-Ministry-of-Health-to-Strengthen-Nationwide-Cardiac-Care), 2026-06-08
- [Deployment] Cabo Verde · Nurse — 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. [balai.cv](https://www.balai.cv/noticias/missoes-medicas-do-projeto-continuidade-dos-eua-reforcam-saude-em-cabo-verde-com-tecnologia-de-ponta/), 2026-09-13
- [Deployment] Cameroon · Clinician / patient — 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. [facebook.com](https://www.facebook.com/MimiMefoInfo/posts/waspito-an-africa-focused-digital-health-platform-founded-in-cameroon-has-launch/1584186856397413/), 2026-06-25
- [Deployment] Cameroon · GBV case worker — 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. [camer.be](https://www.camer.be/93809/11:1/cameroun-wetech-deploie-des-solutions-numeriques-innovantes-pour-la-protection-des-femmes-en-zone-de-crise-cameroon.html), 2026-07-02
- [Regulation] Czechia — 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. [vzp.cz](https://www.vzp.cz/o-nas/aktuality/nove-technologie-v-peci-o-pacienty-i-pro-praci-lekaru-vzp-spousti-rizeny-vstup-inovaci-do-systemu), 2026-04-27
- [Deployment] Brunei · Primary care clinician — 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. [prnewswire.com](https://www.prnewswire.com/apac/news-releases/davos-spotlight--evyd-ministry-of-health-brunei-darussalam-and-novartis-foundation-launch-national-ai-powered-cardiovascular-disease-prevention-initiative-302670872.html), 2026-01-27
- [Deployment] Estonia · Radiologist — 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. [novaator.err.ee](https://novaator.err.ee/1610127190/eestlaste-tehisaru-aitab-radioloogidel-neeruvahki-margata), 2026-09-03
- [Deployment] Estonia · Doctor — 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. [ojs.utlib.ee](https://ojs.utlib.ee/index.php/EA/article/view/29718), 2026-08-23
- [Deployment] Cambodia · Eye-care worker — 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. [iapb.org](https://www.iapb.org/blog/china-cambodia-digital-eye-health-cooperation-takes-off/), 2026-05-29
- [Deployment] Hungary · Pathologist — 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. [mail.egriugyek.hu](https://mail.egriugyek.hu/egeszseg-plusz/gyorsabb-eredmenyek-pontosabb-diagnozis-a-patologian-csucstechnologia-a-markhot-ferenc-oktatokorhazban-video), 2026-04-30
- [Deployment] Hungary · Orthodontist — 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. [semmelweis.hu](https://semmelweis.hu/hirek/2026/08/27/tobb-ido-a-betegre-strukturalt-adatokkal-es-mesterseges-intelligenciaval-fejlesztik-a-fogaszati-ellatast/), 2026-08-27
- [Deployment] Hungary · Neurologist — 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. [semmelweis.hu](https://semmelweis.hu/hirek/2026/02/26/eloremutato-hazai-kutatasi-osszefogas-a-demencia-elleni-kuzdelemben/), 2026-02-26
- [Deployment] China · Interventional cardiologist — 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. [globaltimes.cn](https://www.globaltimes.cn/page/202609/1369905.shtml), 2026-09-06
- [Deployment] Dominica · Haematologist — 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. [varietynewsgroup.com](https://www.varietynewsgroup.com/post/dominica-launches-the-caribbean-s-first-digital-health-app-for-sickle-cell-disease), 2026-06-19
- [Deployment] Fiji · radiologist — 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. [fbcnews.com.fj](https://www.fbcnews.com.fj/news/nausori-health-centre-gets-digital-radiology-upgrade/), 2026-02-08
- [Deployment] Iceland · nursing and care staff — 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. [island.is](https://island.is/en/news/teton-ai-solution-at-moberg), 2026-05-18
- [Deployment] Iceland · doctor — 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. [island.is](https://island.is/en/news/minni-pappirsvinna-laeknis-meiri-timi-med-skjolstaedingi), 2026-03-09
- [Deployment] Georgia · clinicians and hospital staff — 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. [medplum.com](https://www.medplum.com/blog/medimind-case-study), 2026-08-06
- [Deployment] Latvia · radiologist — 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. [ziemelkurzemesslimnica.lv](https://ziemelkurzemesslimnica.lv/zinas/ziemelkurzemes-regionalaja-slimnica-rentgena-diagnostika-izmanto-maksliga-intelekta-risinajumus/), 2026-08-07
- [Evidence of impact] Honduras — 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. [uhsalud.com](https://www.uhsalud.com/index.php/revhispano/article/view/1184), 2026-07-15
- [Deployment] Moldova · Healthcare administrative staff — 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. [radiomoldova.md](https://radiomoldova.md/p/78757), 2026-06-15
- [Deployment] Kyrgyzstan · rehabilitation therapist — 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. [news.cgtn.com](https://news.cgtn.com/news/2026-08-24/China-donated-smart-rehabilitation-device-put-into-use-in-Kyrgyzstan-1PSjg6WbX1e/p.html), 2026-08-24
- [Deployment] Martinique · interventional neuroradiologist — 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. [martinique.franceantilles.fr](https://www.martinique.franceantilles.fr/actualite/sante/avc-le-chum-renforce-la-prise-en-charge-grace-a-lintelligence-artificielle-1069630.php), 2026-02-23
- [Deployment] Macao · employee — 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. [macaudailytimes.com.mo](https://macaudailytimes.com.mo/tcm-ai-health-station-launched-to-support-worker-well-being.html), 2026-08-12
- [Deployment] Russia · surgeon — 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. [cito-priorov.ru](https://www.cito-priorov.ru/news/145594/), 2026-05-29
- [Deployment] Russia · radiologist — 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. [nn-news.net](https://nn-news.net/society/2026/08/26/891834.html), 2026-08-26
- [Evidence of impact] Malawi — 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. [english.elpais.com](https://english.elpais.com/health/2026-06-11/hospitals-in-malawi-team-up-with-artificial-intelligence-to-reduce-child-mortality.html), 2026-06-11
- [Deployment] Namibia · hospital administrative staff, nursing staff — 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. [proto.cx](https://proto.cx/resource/medfam-hospital-network-deploys-proto-ai-patient-experience-agents/), 2026-04-01
- [Deployment] Slovakia · neurosurgeon — 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. [medicina.trend.sk](https://medicina.trend.sk/inovacie/neurochirurgovia-z-kramarov-maju-za-sebou-unikatnu-operaciu-mozgu-asistovala-im-umela-inteligencia/), 2026-01-25
- [Deployment] Myanmar · ophthalmologist — 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. [asianews.network](https://asianews.network/ophthalmology-hospital-in-yangon-conducts-field-eye-examinations-using-ai/), 2026-05-26
- [Deployment] Estonia · Mental health counsellor — 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. [mu.ee](https://www.mu.ee/uudised/2026/09/03/peaasiee-votab-soome-eeskujul-kasutusele-turvalise-noustamisroboti), 3 Sep 2026
- [Deployment] Finland · Social and health care professional — 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. [sttinfo.fi](https://www.sttinfo.fi/tiedote/72254203/lansi-uudenmaan-hyvinvointialue-uudistaa-kirjaamista-tekoalyn-avulla?publisherId=69818831&lang=fi), 24 Aug 2026
- [Deployment] United Kingdom · radiologist — 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. [gov.uk](https://www.gov.uk/government/news/ai-to-speed-up-cancer-diagnosis-for-millions-of-nhs-patients), 2026-06-10
- [Regulation] United Kingdom — 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. [gov.uk](https://www.gov.uk/government/news/independent-commission-led-by-nhs-doctors-sets-out-blueprint-to-accelerate-safe-ai-adoption-in-healthcare), 2026-09-10
- [Deployment] North Korea · Korean medicine doctors — 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. [nkeconomy.com](http://www.nkeconomy.com/news/articleView.html?idxno=16645), 2026-05-27
- [Deployment] Papua New Guinea · radiographer / community health worker — 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. [thenational.com.pg](https://www.thenational.com.pg/ai-technology-combating-tb/), 2026-04-02
- [Deployment] Turks and Caicos Islands · healthcare case manager — 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. [newslinetci.com](https://www.newslinetci.com/post/saunders-pitches-smart-safe-city-digital-transformation-in-budget-debate), 2026-04-30
- [Deployment] Venezuela · surgeon — 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. [noticierove.com](https://noticierove.com/inauguran-servicios-de-cirugia-robotica-y-de-cirugia-plastica-reconstructiva-en-el-huc/), 2026-05-26
- [Deployment] Sierra Leone · public health officials — 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. [theglobalpressnetwork.com](https://theglobalpressnetwork.com/2026/06/25/sierra-leone-tests-ai-powered-smart-trap-to-track-malaria-mosquitoes/), 2026-06-25
- [Deployment] Botswana · wellness programme manager — 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. [einpresswire.com](https://www.einpresswire.com/article/904505886/bomaid-launches-thrive-with-bomaid-powered-by-vivascore), 2026-04-08
- [Deployment] Jamaica · Radiologist — 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. [jamaica-gleaner.com](https://jamaica-gleaner.com/article/news/20260801/tufton-hails-ai-powered-imaging-centre), 2026-08-01
- [Deployment] Malta · Doctor — 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. [maltaceos.mt](https://maltaceos.mt/word-of-mouth-is-still-our-strongest-form-of-marketing-saint-james-hospital-ceo/), 2026-03-29
- [Deployment] Saint Kitts and Nevis · clinician — 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. [marketscale.com](https://www.marketscale.com/industries/healthcare/st-kitts-and-nevis-launches-national-digital-health-platform-covering-51000-citizens), 2026-07-19
- [Deployment] Suriname · health educator — 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. [surinamenieuwscentrale.com](https://www.surinamenieuwscentrale.com/ai-assistent-flowvie-moet-taboe-rond-menstruatie-bij-jonge-vrouwen-doorbreken), 2026-05-26
- [Deployment] United Kingdom · Operating theatre team — 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. [press.aboutamazon.com](https://press.aboutamazon.com/2026/9/global-healthtech-proximie-in-collaboration-with-aws-deloitte-and-the-nhs-launches-largest-nhs-ai-theatre-evaluation-to-unlock-hidden-operating-theatre-capacity), 2026-09-09
- [Regulation] United Kingdom — 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. [gov.uk](https://www.gov.uk/government/news/mhra-and-manchester-nhs-partner-on-health-innovation-sandbox), 2026-09-02
- [Deployment] Spain · Primary care doctor — 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. [espaciodiario.com](https://espaciodiario.com/la-comunidad-desplegara-proyectos-con-ia-para-anticipar-enfermedades-cronicas-y-personalizar-la-atencion-sanitaria/), 2026-09-10
- [Deployment] Nicaragua · Radiologists and pathologists — 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. [canal2tv.com](https://canal2tv.com/nacionales/centro-oncologia-ia-tecnologia/), 2026-06-24
- [Deployment] Canada · radiologist / cardiologist — 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. [thewhig.com](https://www.thewhig.com/news/kingston-hospital-first-in-canada-to-adopt-ai-tool-for-heart-disease-diagnosis), 2025-08-14
- [Deployment] Canada · medical secretary / triage nurse — 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. [ici.radio-canada.ca](https://ici.radio-canada.ca/nouvelle/2152448/navig-intelligence-artificielle-sante-quebec), 2025-04-02
- [Deployment] Canada · nurse / care aide — 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. [esprit-ai.com](https://esprit-ai.com/from-pilot-to-practice-bruyere-health-officially-adopts-esprit-ais-ai-solution-to-prevent-falls-and-pressure-injuries-in-long-term-care/), 2025-07-15

## Legal & Professional Services

- [Deployment] United States · Lawyer — 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. [thomsonreuters.com](https://thomsonreuters.com/en/press-releases/2026/august/greenberg-traurig-deploys-the-next-generation-of-cocounsel-legal), 26 Aug 2026
- [Deployment] United States · Lawyer — 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. [googlecloudpresscorner.com](https://googlecloudpresscorner.com/2026-08-25-Google-Cloud-Launches-Gemini-Enterprise-for-Legal), 25 Aug 2026
- [Deployment] United States · Lawyer — 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. [ballardspahr.com](https://ballardspahr.com/insights/news/2026/08/ballard-spahr-expands-ai-adoption-with-firm-wide-rollout-of-syllo-litigation-agent), 5 Aug 2026
- [Deployment] United States · Lawyer — 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. [faegredrinker.com](https://faegredrinker.com/en/insights/publications/2026/8/scott-angelo-discusses-firmwide-ai-adoption-with-law-week-colorado), 13 Aug 2026
- [Deployment] United States · Lawyer — 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. [bakerlaw.com](https://www.bakerlaw.com/insights/bakerhostetler-unveils-practice-intelligence-center-and-strategic-collaboration-with-clio/), 26 Aug 2026
- [Deployment] United States · Lawyer — 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. [nelsonmullins.com](https://www.nelsonmullins.com/insights/news/press_releases/nelson-mullins-deploys-harvey-across-all-practices), 24 Aug 2026
- [Deployment] United States · Lawyer — 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. [dwt.com](https://www.dwt.com/about/news/2026/08/dwt-expands-firmwide-ai-capabilities-with-harvey), 18 Aug 2026
- [Deployment] United States · Lawyer — 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. [afslaw.com](https://www.afslaw.com/perspectives/press-releases/arentfox-schiff-introduces-foxai-client-first-ai-innovation-framework), 17 Aug 2026
- [Deployment] United States · Lawyer — 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. [abajournal.com](https://www.abajournal.com/news/article/goodwin-procter-unveils-ai-enhanced-tool-for-venture-capital-funding), 14 Aug 2026
- [Deployment] Netherlands · Consultant — Haute Equipe: Haute Equipe has rolled out soev.ai, a Dutch sovereign-AI platform, across the firm. Every employee got AI-literacy training. [consultancy.nl](https://www.consultancy.nl/nieuws/69362/), 2026-08-07
- [Deployment] Czechia · Lawyer — 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. [legaltechnology.com](https://legaltechnology.com/kinstellar-selects-laurel-ai-in-commercial-excellence-drive/), 2026-08-20
- [Deployment] United Kingdom · Lawyer — 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. [legaltechnology.com](https://legaltechnology.com/from-the-orange-rag-clifford-chance-reimagines-knowledge-management-for-the-ai-era/), 2026-08-21
- [Deployment] Germany · Tax / legal advisor — 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. [silicon.de](https://www.silicon.de/41723677/40-000-stunden-mehr-zeit-fuer-beratung), 2026-08-30
- [Deployment] South Africa · Consultant / auditor — 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. [cnbcafrica.com](https://www.cnbcafrica.com/2026/from-ambition-to-operating-reality-ai-summit-shows-african-business-already-building-the-future), 2026-08-28
- [Evidence of adoption] Global — 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. [thomsonreuters.com](https://www.thomsonreuters.com/en-us/posts/technology/agentic-ai-oversight-challenges/), 2025
- [Evidence of adoption] United States — 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. [americanbar.org](https://www.americanbar.org/groups/law_practice/resources/tech-report/2024/2024-artificial-intelligence-techreport/), 2024
- [Evidence of adoption] United Kingdom — By September 2025, 61% of UK lawyers used generative AI in day-to-day work. In January, AI use stood at 46%. [lexisnexis.com](https://www.lexisnexis.com/community/pressroom/b/news/posts/two-thirds-of-uk-lawyers-now-use-ai-yet-firm-culture-slows-progress), 2025
- [Deployment] Global · Lawyer — 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. [ashurstperkinscoie.com](https://www.ashurstperkinscoie.com/en/insights/vox-populai-lessons-from-a-global-law-firms-exploration-of-generative-ai/), 2024
- [Evidence of adoption] Global — 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. [prnewswire.com](https://www.prnewswire.com/news-releases/anti-money-laundering-pros-find-expanding-uses-for-ai-but-adoption-remains-slow-302384474.html), 
- [Evidence of adoption] Global — 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. [thomsonreuters.com](https://www.thomsonreuters.com/en/institute/articles/ai-in-professional-services-report-2026), 2026
- [Deployment] United Kingdom · Compliance officer — 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. [hsbc.com](https://www.hsbc.com/news-and-views/views/hsbc-views/harnessing-the-power-of-ai-to-fight-financial-crime), 
- [Evidence of impact] Global — 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. [damiencharlotin.com](https://www.damiencharlotin.com/hallucinations/), 2026
- [Deployment] United States · Lawyer — 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. [lawnext.com](https://www.lawnext.com/2026/03/sixth-circuit-slaps-steep-sanctions-on-two-lawyers-for-fake-citations-and-misrepresentations-in-appellate-briefs.html), 2026
- [Evidence of adoption] Global — 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. [thomsonreuters.com](https://www.thomsonreuters.com/en/institute/articles/ai-in-professional-services-report-2026), 2026
- [Evidence of impact] United States — 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. [thomsonreuters.com](https://www.thomsonreuters.com/en/institute/articles/ai-in-professional-services-report-2026), 2026
- [Regulation] United States — 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. [mprnews.org](https://www.mprnews.org/story/2026/08/21/lawyer-in-minnesota-suspended-over-fake-ai-case-citations), 2026-08-21
- [Deployment] Mexico · In-house litigation lawyer — 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. [harvey.ai](https://www.harvey.ai/blog/grupo-financiero-inbursa-deploys-harvey), 2026-09-07
- [Deployment] South Africa · Lawyer — 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. [legora.com](https://legora.com/newsroom/bowmans-rolls-out-legora-to-enhance-legal-delivery), 2026-03-03
- [Deployment] Ghana · Government lawyer — 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. [techlabari.com](https://techlabari.com/ghanas-finance-ministry-turns-to-eskwai-for-legal-research/), 2026-08-26
- [Regulation] Colombia — 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. [infobae.com](https://www.infobae.com/colombia/2026/02/16/la-corte-suprema-sanciona-por-primera-vez-a-un-abogado-por-uso-indebido-de-inteligencia-artificial-en-proceso-judicial/), 2026-02-16
- [Regulation] Colombia — 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. [eltiempo.com](https://www.eltiempo.com/justicia/cortes/el-duro-llamado-de-atencion-del-consejo-de-estado-sobre-el-uso-irresponsable-de-la-inteligencia-artificial-en-demandas-3561102), 2026-06-01
- [Deployment] Argentina · Lawyer — 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. [eldestapeweb.com](https://www.eldestapeweb.com/politica/inteligencia-artificial/la-corte-suprema-suma-inteligencia-artificial-para-los-fallos-judiciales-20263317310), 2026-03-03
- [Regulation] Chile — 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. [adprensa.cl](https://adprensa.cl/corte-suprema-sanciona-a-abogada-por-citas-doctrinales-inexistentes-generadas-por-inteligencia-artificial/), 2026-06-03
- [Deployment] Brazil · Public defender — 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. [defensoria.rj.def.br](https://defensoria.rj.def.br/noticia/detalhes/32480-DPRJ-aposta-em-Inteligencia-Artificial-para-acelerar-em-ate-5x-o-tempo-de-atendimento-a-populacao), 2026-03-06
- [Evidence of adoption] Brazil — 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. [direitonews.com.br](https://www.direitonews.com.br/2026/04/ia-ja-usada-77-advogados-redefine-pratica-juridica-aponta-estudo-oab.html), 2026-04-01
- [Deployment] Brazil · Lawyer — 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. [debatejuridico.com.br](https://debatejuridico.com.br/oab-lanca-opendetector/), 2026-07-13
- [Regulation] Qatar — 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. [pinsentmasons.com](https://www.pinsentmasons.com/out-law/news/qatar-introduces-new-ai-usage-rules-for-courts), 2026-02-10
- [Evidence of impact] United States — 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. [abajournal.com](https://www.abajournal.com/news/article/dc-circuit-blames-deutsche-bank-lawyers-for-ai-hallucinations), 2026-09-04
- [Regulation] Lebanon — 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. [albaladnews.net](https://albaladnews.net/article/473242), 2026-09-10
- [Deployment] Azerbaijan · Architect — 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. [vesti.az](https://vesti.az/texnologiya/tureckaya-kompaniya-ispolzovala-ii-v-stroitelnyx-proektax-v-azerbaidzane-593678), 2026-05-11
- [Evidence of impact] Estonia — 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. [err.ee](https://www.err.ee/1610007988/tehisintellekt-on-votnud-tolkijatelt-too), 2026-04-27
- [Deployment] Hungary · Lawyer — 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. [kormany.hu](https://kormany.hu/hirek/tuzson-bence-elindult-a-mestersegesintelligencia-alapu-jogi-informaciokereso), 2026-03-16
- [Deployment] Hong Kong · Lawyer — 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. [conventuslaw.com](https://conventuslaw.com/press-releases/hong-kong-johnson-stokes-master-advances-governance-first-human-led-ai-to-support-legal-service-delivery-with-microsoft-copilot-agents/), 2026-07-13
- [Deployment] Liberia · citizen — 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. [thenewdawnliberia.com](https://www.thenewdawnliberia.com/liblaw-launches-liberias-first-legal-learning-artificial-intelligence/), 2026-08-03
- [Regulation] Nepal — 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. [kathmandupost.com](https://kathmandupost.com/national/2026/07/03/nepal-s-courts-embrace-ai-keep-judgment-writing-human), 2026-07-03
- [Regulation] Guam — 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. [postguam.com](https://www.postguam.com/news/local/guam-supreme-court-sanctions-phillips-on-ai-issue/article_e3840321-95d9-46dc-a97f-04f8b4ea1334.html), 2026-07-12
- [Regulation] Guam — 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. [postguam.com](https://www.postguam.com/news/local/another-lawyer-gets-sanctions-over-ai-issue/article_9f2c13c2-9e37-4d7b-a35a-a52de809bf5c.html), 2026-08-02
- [Deployment] Guadeloupe · lawyer — 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. [rci.fm](https://rci.fm/guadeloupe/infos/Justice/Lintelligence-artificielle-integre-certains-cabinets-davocats), 2026-01-17
- [Deployment] Canada · lawyer — 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. [gowlingwlg.com](https://gowlingwlg.com/en-ca/news/firm-news/2025/gowling-wlg-becomes-first-canadian-law-firm-to-roll-out-harvey-ai-enterprise-wide), 2025-07-11
- [Deployment] Canada · lawyer — 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. [blg.com](https://www.blg.com/en/about-us/news/2025/12/blg-expands-long-standing-thomson-reuters-partnership-with-firmwide-rollout-of-cocounsel), 2025-12-03
- [Deployment] Canada · lawyer — 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. [newswire.ca](https://www.newswire.ca/news-releases/mcmillan-unveils-canadian-flagship-partnership-with-legora-to-enhance-client-service-delivery-804257221.html), 2025-12-17
- [Deployment] Canada · lawyer — 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. [alexi.com](https://www.alexi.com/blog/kmsc-law-deploys-alexis-legal-intelligence-platform-in-private-cloud-to-scale-ai-firm-wide), 2025-07-31

## Finance & Insurance

- [Deployment] United Kingdom · Treasury / FX banker — 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. [globalbankingandfinance.com](https://www.globalbankingandfinance.com/citi-hsbc-stanchart-adopt-ant-international-forex-ai-tool/), 20 Aug 2026
- [Deployment] Germany · Financial analyst — 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. [googlecloudpresscorner.com](https://www.googlecloudpresscorner.com/2026-08-25-Google-Cloud-Launches-Gemini-Enterprise-for-Financial-Services), 25 Aug 2026
- [Deployment] United States · Financial analyst — 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. [googlecloudpresscorner.com](https://www.googlecloudpresscorner.com/2026-08-25-Google-Cloud-Launches-Gemini-Enterprise-for-Financial-Services), 25 Aug 2026
- [Deployment] United States · Banking staff — 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. [googlecloudpresscorner.com](https://www.googlecloudpresscorner.com/2026-08-25-Google-Cloud-Launches-Gemini-Enterprise-for-Financial-Services), 25 Aug 2026
- [Deployment] United States · Consumer-facing — 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. [pymnts.com](https://pymnts.com/news/artificial-intelligence/2026/visa-readies-rollout-ai-financial-assistant-banking-apps/), pilot begins Aug 2026
- [Deployment] Singapore · Corporate banker — 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. [dbs.com](https://www.dbs.com/newsroom/DBS_scales_agentic_AI_to_transform_way_of_working_for_corporate_bankers_freeing_up_time_for_more_strategic_client_engagements), 19 Aug 2026
- [Deployment] South Korea · Customer-service agent — 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. [en.sedaily.com](https://en.sedaily.com/technology/2026/08/31/kt-to-rebuild-woori-banks-ai-chatbot-and-consultation-bot), 31 Aug 2026
- [Deployment] India · Underwriter / claims — 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. [salesforce.com](https://www.salesforce.com/in/news/press-releases/2026/08/27/indiafirst-lifes-ai-driven-transformation-with-salesforce-agentforce/), 2026-08-27
- [Deployment] Netherlands · Customer service / compliance — 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. [computable.nl](https://www.computable.nl/2026/08/13/abn-amro-voortvarend-met-ai/), 2026-08-13
- [Deployment] United Kingdom · Consumer-facing — 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. [experianplc.com](https://www.experianplc.com/newsroom/press-releases), 2026-08-27
- [Deployment] United Kingdom · Consumer-facing — 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. [experianplc.com](https://www.experianplc.com/newsroom/press-releases), 2026-08-20
- [Deployment] Canada · Banker / staff — 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. [cibc.mediaroom.com](https://cibc.mediaroom.com/2026-07-31-CIBC-introduces-the-first-enterprise-wide-agentic-AI-workspace-in-Canadian-banking), 2026-07-31
- [Deployment] Brazil · Consumer-facing — 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. [riotimesonline.com](https://www.riotimesonline.com/itau-genai-assistant-iai-launch-2026/), 2026-07-27
- [Deployment] United Arab Emirates · Banking operations — 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. [khaleejtimes.com](https://www.khaleejtimes.com/business/uae-digital-bank-ruya-deploys-agentic-ai-targets-ai-native-future), 2026-08-18
- [Deployment] South Africa · Customer service — 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. [itweb.co.za](https://www.itweb.co.za/article/absas-ai-chatbot-reaches-16m-users/JBwErvn3WQd76Db2), 2026-08-20
- [Deployment] South Africa · Customer service — 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. [cnbcafrica.com](https://www.cnbcafrica.com/2026/from-ambition-to-operating-reality-ai-summit-shows-african-business-already-building-the-future), 2026-08-28
- [Evidence of adoption] United Kingdom — 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. [bankofengland.co.uk](https://www.bankofengland.co.uk/report/2024/artificial-intelligence-in-uk-financial-services-2024), 2024
- [Evidence of impact] United States — 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. [home.treasury.gov](https://home.treasury.gov/news/press-releases/jy2650), 2024
- [Deployment] United States · Banker — 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. [jpmorganchase.com](https://www.jpmorganchase.com/about/technology/blog/llmsuite-ab-award), 2025
- [Deployment] Australia · Fraud analyst — 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. [commbank.com.au](https://www.commbank.com.au/articles/newsroom/2026/04/ai-agent-spots-fraud-in-real-time.html), 2026
- [Deployment] United Kingdom · Regulator — 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. [fca.org.uk](https://www.fca.org.uk/firms/innovation/ai-lab), 2026
- [Evidence of impact] United States — 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. [fiscal.treasury.gov](https://fiscal.treasury.gov/dnp/), 2026
- [Deployment] Colombia · retail banking staff — 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. [semana.com](https://www.semana.com/foros-semana/articulo/asi-es-como-la-banca-esta-dejando-atras-los-pilotos-con-inteligencia-artificial-y-la-esta-integrando-a-su-negocio/202608/), 2026-08-31
- [Regulation] Bermuda — 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. [bernews.com](https://bernews.com/2026/08/bma-issues-consultation-paper-on-ai-use/), 2026-08-15
- [Evidence of adoption] Argentina — 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. [anda.cl](https://anda.cl/trafico-referido-ia-banca-latinoamerica/), 2026-08-27
- [Deployment] Peru · corporate banker — 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. [forbes.pe](https://forbes.pe/tecnologia/ia/2026-08-11/interbank-suscribio-acuerdo-con-anthropic-para-fortalecer-el-uso-de-inteligencia-artificial-en-el-banco/), 2026-08-11
- [Deployment] United States · banker — 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. [artificialintelligence-news.com](https://www.artificialintelligence-news.com/news/mt-bank-enterprise-ai-15000-employees/), 2026-09-04
- [Evidence of impact] United States — 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. [claimsjournal.com](https://www.claimsjournal.com/news/national/2026/08/27/339781.htm), 2026-08-27
- [Deployment] Canada · financial advisor — 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. [ca.finance.yahoo.com](https://ca.finance.yahoo.com/news/sun-life-puts-ai-advisors-140000551.html), 2026-08-19
- [Regulation] Hong Kong — 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. [thestandard.com.hk](https://www.thestandard.com.hk/innovation/article/341121/Hong-Kong-unveils-first-cohort-of-GenAI-Sandbox), 2026-08-27
- [Deployment] South Korea · Complaints counsellor — 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. [en.sedaily.com](https://en.sedaily.com/finance/2026/09/07/korea-life-insurance-association-adopts-ai-for-complaint), 2026-09-07
- [Deployment] Singapore · Banker — 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. [qa-financial.com](https://qa-financial.com/uob-puts-testing-behind-300-ai-deployments/), 2026-09-01
- [Deployment] Uzbekistan · Credit officer — 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. [trend.az](https://www.trend.az/casia/uzbekistan/4219325.html), 2026-09-01
- [Deployment] Australia · Business banker — 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. [anz.com.au](https://www.anz.com.au/newsroom/media/2026/february/anz-launches-agentic-ai-powered-crm-to-transform-business-banking/), 2026-02-05
- [Deployment] Indonesia · Relationship manager — 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. [theasianbanker.com](https://www.theasianbanker.com/press-releases/cimb-niaga-partners-with-google-cloud-and-artefact-to-deploy-ai-agents-for-banking-staff-in-indonesia), 2026-04-14
- [Evidence of adoption] China — 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. [finance.sina.com.cn](https://finance.sina.com.cn/wm/2026-04-17/doc-inhuvres2073569.shtml), 2026-04-17
- [Deployment] Canada · Insurance underwriter — 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. [newswire.ca](https://www.newswire.ca/news-releases/manulife-canada-delivers-faster-life-insurance-approvals-with-ai-integration-and-enhanced-digital-application-898645145.html), 2026-01-29
- [Deployment] Canada · Contact-centre adviser — 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. [newswire.ca](https://www.newswire.ca/news-releases/scotiabank-launches-scotia-intelligence-empowering-employees-and-accelerating-enterprise-ai-adoption-854931680.html), 2026-04-13
- [Deployment] Japan · Contact centre operator — 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. [pkshatech.com](https://www.pkshatech.com/news/20260304/), 2026-03-04
- [Deployment] Japan · Call centre operator — 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. [jp.newsroom.ibm.com](https://jp.newsroom.ibm.com/2026-02-18-SMBC-AI-Operator), 2026-02-18
- [Deployment] South Korea · Central bank analyst — 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. [seoul.co.kr](https://www.seoul.co.kr/news/economy/2026/01/25/20260125500031), 2026-01-25
- [Regulation] China — 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. [21jingji.com](https://www.21jingji.com/article/20260624/7e99cde3619cf5a7ca885e5c18e68d78.html), 2026-06-24
- [Deployment] Ghana · Central bank economist — 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. [thebftonline.com](https://thebftonline.com/article/bog-deploys-ai-and-big-data-to-speed-up-economic-intelligence), 2026-08-27
- [Deployment] Colombia · Corporate relationship manager — 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. [bbva.com](https://www.bbva.com/es/empresas/bbva-lanza-brujula-para-transformar-con-inteligencia-artificial-el-asesoramiento-a-las-empresas-en-america-del-sur/), 2026-08-27
- [Deployment] Portugal · In-house lawyer — 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. [jornaleconomico.sapo.pt](https://jornaleconomico.sapo.pt/noticias/bpi-adota-plataforma-de-ia-harvey-para-reforcar-eficiencia-do-departamento-juridico/), 2026-06-22
- [Evidence of adoption] Belgium — 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. [fintechbelgium.be](https://www.fintechbelgium.be/news/ai-barometer-report-2026), 2026-06-18
- [Deployment] Mexico · Bank operations analyst — 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. [ir.uipath.com](https://ir.uipath.com/news/detail/459/banco-azteca-evolves-from-automation-to-enterprise-scale-business-orchestration-with-uipath), 2026-08-26
- [Deployment] Kenya · Motor claims assessor — 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. [techmoran.com](https://techmoran.com/2026/02/27/britam-launches-ai-service-to-settle-motor-claims-within-two-hours/), 2026-02-27
- [Regulation] Nigeria — 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. [techcabal.com](https://techcabal.com/2026/03/12/cbn-wants-ai-to-fight-money-laundering/), 2026-03-12
- [Deployment] South Africa · Bank client-support agent — 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. [techcentral.co.za](https://techcentral.co.za/capitec-bets-big-on-ai-and-keeps-hiring/280534/), 2026-04-22
- [Deployment] Nigeria · Insurance customer-service agent — 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. [heirsinsurancegroup.com](https://www.heirsinsurancegroup.com/heirs-insurance-leads-africas-insurance-revolution-with-nigerias-first-multi-language-genai-assistant/), 2026-05-12
- [Deployment] South Africa · Credit underwriter — 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. [techcentral.co.za](https://techcentral.co.za/nedbank-jumo-bet-on-ai-lending-for-the-underbanked/282272/), 2026-06-04
- [Deployment] Dominican Republic · Contact-centre agent — 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. [diariolibre.com](https://www.diariolibre.com/economia/negocios/2026/07/15/el-banco-popular-utiliza-ia-para-coordinar-entrega-de-tarjetas/3600400), 2026-07-15
- [Evidence of impact] Canada — 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. [insurancejournal.com](https://www.insurancejournal.com/services/newswire/2026/09/10/884569.htm), 2026-09-10
- [Deployment] Canada · Mortgage underwriter — 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. [td.mediaroom.com](https://td.mediaroom.com/2026-05-21-TD-Launches-Agentic-AI-to-Transform-Real-Estate-Secured-Lending-from-End-to-End), 2026-05-21
- [Deployment] Azerbaijan · Retail banking customer service agent — 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. [aap.com.au](https://www.aap.com.au/aapreleases/globenewswire1001185827/), 2026-06-04
- [Deployment] Azerbaijan · Payments customer service agent — 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. [infocity.tech](https://infocity.tech/2026/08/bir-zapuskaet-pervogo-v-azerbajdzhane-ii-assistenta-dlya-elektronnyh-koshelkov/), 2026-08-11
- [Deployment] Cameroon · Insurance broker — 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. [vitrineducameroun.com](https://vitrineducameroun.com/2026/02/27/cova-lance-une-solution-basee-sur-lintelligence-artificielle-et-lagregation-de-paiement-mobile-pour-simplifier-lacces-a-lassurance-automobile-obligatoire-au-cameroun/), 2026-03-02
- [Deployment] Czechia · Financial licensing specialist — 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. [cnb.cz](https://www.cnb.cz/cs/cnb-news/tiskove-zpravy/CNB-vydala-prvnich-sest-povoleni-podle-narizeni-MiCA-00001/), 2026-02-11
- [Deployment] Czechia · Internal auditor — 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. [ekonom.cz](https://ekonom.cz/c1-67843960-kryptomeny-uz-nejsou-pro-banky-tabu-raiffeisenbank-jako-prvni-velka-banka-v-cesku-je-chce-letos-dostat-do-sve-bankovni-aplikace), 2026-02-11
- [Evidence of impact] Czechia — 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. [lupa.cz](https://www.lupa.cz/aktuality/cnb-zalozila-sekci-pro-ai-vlivem-nasazovani-umele-inteligence-uz-se-take-propoustelo/), 2026-07-31
- [Regulation] Guernsey — 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. [gfsc.gg](https://www.gfsc.gg/news/policy-statement-use-artificial-intelligence), 2026-01-29
- [Deployment] DR Congo · Entrepreneur — 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. [actualite.cd](https://actualite.cd/2026/02/04/entrepreneuriat-en-rdc-bokeli-la-plateforme-digitale-intelligente-du-fogec-pour-la), 2026-02-04
- [Deployment] Hungary · Claims handler — 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. [vg.hu](https://www.vg.hu/vilaggazdasag-magyar-gazdasag/2026/04/biztositas-uniqa-mesterseges-intelligencia-csotores), 2026-04-20
- [Evidence of adoption] Hungary — 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. [economx.hu](https://www.economx.hu/magyar-gazdasag/2026/09/08/ai-forradalom-a-magyar-bankokban-170-helyen-mar-elesben-dolgozik-a-mesterseges-intelligencia/), 2026-09-08
- [Deployment] Isle of Man · Claims handler — 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. [investegate.co.uk](https://www.investegate.co.uk/announcement/rns/manx-financial-group--mfx/h1-fy-2026-operational-update-/9755145), 2026-09-04
- [Deployment] Georgia · data analysts and banking staff — 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. [contextwindows.ai](https://www.contextwindows.ai/case-study/tbc-bank-business-analytics), 2026-06-09
- [Deployment] Latvia · customer service agent — 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. [seb.lv](https://www.seb.lv/info/jaunumi/seb-banka-ievies-balss-robotu-steidzamu-palidzibu-klienti-vares-sanemt-vel-atrak), 2026-08-11
- [Deployment] Kyrgyzstan · bank customer service agent — 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. [economist.kg](https://economist.kg/novosti-kompanii/2026/08/12/finca-ai-bot-kyrgyz-language/), 2026-08-12
- [Evidence of impact] Russia — 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. [vedomosti.ru](https://www.vedomosti.ru/finance/news/2026/06/30/1209942-sber-effekt-ii), 2026-06-30
- [Deployment] Maldives · Customer service representative — 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. [en.mmtv.mv](https://en.mmtv.mv/10958), 2026-07-22
- [Deployment] Serbia · Central bank economist — 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. [srpske.rs](https://srpske.rs/en/news/ekonomija/2026/08/20/nbs-ai-model-daily-price-lists-september), 2026-08-20
- [Evidence of adoption] Serbia — 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. [kpmg.com](https://kpmg.com/rs/en/insights/consulting/2026/08/ai-in-the-risk-function-of-serbian-banks.html), 2026-08-17
- [Deployment] Myanmar · insurance agent — 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. [ayasompo.com](https://ayasompo.com/introducing-iva-bot/), 2026-04-29
- [Deployment] Switzerland · underwriter — 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. [financialit.net](https://financialit.net/news/insurtech/zurich-scales-agentic-ai-five-countries-90-days-cutting-manual-risk-processing-time), 2026-05-18
- [Deployment] United Kingdom · personal banking customer — 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. [starlingbank.com](https://www.starlingbank.com/news/starling-launches-agentic-ai-assistant/), 2026-03-20
- [Deployment] United Kingdom · finance professional, engineer — 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. [technologyrecord.com](https://www.technologyrecord.com/article/nationwide-deploys-microsoft-365-copilot-to-modernise-banking-behind-the-scenes), 2026-09-08
- [Regulation] Jersey — 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. [mourant.com](https://www.mourant.com/updates/jfsc-ai-jersey/), 2026-07-17
- [Deployment] Monaco · private banker — 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. [finnoex.com](https://finnoex.com/cmb-monaco-becomes-first-to-go-live-on-integrated-avaloq-and-aladdin-wealth-platform/), 2026-04-02
- [Deployment] Eswatini · treasury professional — 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. [times.co.sz](https://times.co.sz/40957/business/sbs-bank-goes-live-with-emach-ai-treasury-platform/), 2026-08-06
- [Evidence of adoption] Switzerland — 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. [the-decoder.com](https://the-decoder.com/at-ubs-ai-skills-are-now-a-condition-for-landing-a-job/), 2026-09-07
- [Deployment] Canada · Credit adjudicator, risk manager — 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. [newswire.ca](https://www.newswire.ca/news-releases/rbc-s-proprietary-ai-foundation-model-for-financial-services-aims-to-enhance-personalized-client-experience-846568882.html), 2025-07-29
- [Deployment] Canada · Securities professional, trader, research analyst — 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. [newswire.ca](https://www.newswire.ca/news-releases/virtual-ai-assistant-to-help-power-td-securities-813020004.html), 2025-07-08
- [Deployment] Canada · Banking professional, marketing manager — 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. [newswire.ca](https://www.newswire.ca/news-releases/cibc-launches-cibc-crtex-tm-an-ai-enabled-client-personalization-amp-engagement-engine-to-further-its-industry-leading-digital-capabilities-and-enhance-client-experiences-890473929.html), 2025-10-16
- [Deployment] Canada · Underwriter — 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. [insurance-canada.ca](https://insurance-canada.ca/2025/05/30/markel-expands-mea-agentic-ai-platform/), 2025-05-30
- [Deployment] Canada · Underwriter — 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. [insurance-edge.net](https://insurance-edge.net/2025/11/19/cna-canada-and-mea-partner-up-on-ai-solution/), 2025-11-19

## Education & Training

- [Deployment] United States · Teacher — 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. [edtechinnovationhub.com](https://edtechinnovationhub.com/news/openai-expands-chatgpt-for-teachers-to-55-more-us-school-systems), 26 Aug 2026
- [Deployment] United States · Teacher — 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. [live5news.com](https://www.live5news.com/2026/08/11/students-return-school-ai-takes-seat-charleston-county-classrooms/), 11 Aug 2026
- [Deployment] United States · Teacher — 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. [districtadministration.com](https://districtadministration.com/article/in-this-district-ai-is-a-teachers-personal-assistant/), 20 Aug 2026
- [Deployment] United States · Teacher — 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. [wect.com](https://www.wect.com/2026/08/18/ai-classroom-how-local-school-districts-are-managing-tech-expansion/), 18 Aug 2026
- [Deployment] United States · Teacher — 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. [wect.com](https://www.wect.com/2026/08/18/ai-classroom-how-local-school-districts-are-managing-tech-expansion/), 18 Aug 2026
- [Deployment] United States · Faculty / student — 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. [oit.duke.edu](https://oit.duke.edu/news/updated-duke-ai-tools-and-upcoming-webinars-fall-2026/), 12 Aug 2026
- [Deployment] United States · Faculty / student — 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. [oit.colorado.edu](https://oit.colorado.edu/news/chatgpt-edu-now-available-students-new-training-courses-are-live), 14 Aug 2026 go-live
- [Deployment] Singapore · Faculty / student — 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. [ntu.edu.sg](https://ntu.edu.sg/news/detail/empowering-students-to-thrive-in-the-age-of-ai), Aug 2026 go-live
- [Deployment] Singapore · Faculty / student — 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. [news.nus.edu.sg](https://news.nus.edu.sg/nus-powers-education-research-and-administration-to-new-heights-with-ai-through-a-strategic-collaboration-with-openai/), 2026-08-11
- [Deployment] Australia · Faculty / staff — 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. [westernsydney.edu.au](https://www.westernsydney.edu.au/news-centre/stories/2026/western-sydney-university-embraces-microsoft-copilot-for-all-staff), 2026-07-22
- [Deployment] Austria · Student — 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. [aau.at](https://www.aau.at/en/blog/academic-ai-available-to-students-from-the-winter-semester/), 2026-09-01
- [Deployment] Denmark · Faculty / staff — 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. [via.ritzau.dk](https://via.ritzau.dk/pressemeddelelse/15115214/), 2026-08-27
- [Deployment] Italy · Faculty / student — 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. [univrmagazine.it](https://univrmagazine.it/2026/07/30/), 2026-07-30
- [Evidence of adoption] United States — AI chatbots help many U.S. teens with schoolwork. By late 2025, 54% of 13–17-year-olds said they had used one. [pewresearch.org](https://www.pewresearch.org/internet/2026/02/24/how-teens-use-and-view-ai/), 2025
- [Evidence of adoption] United States — 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. [news.gallup.com](https://news.gallup.com/poll/691967/three-teachers-weekly-saving-six-weeks-year.aspx), 
- [Evidence of adoption] United Kingdom — In 2026, 95% of UK undergraduates report some AI use and 94% use generative AI for assessed work. Overall use was 66% in 2024. [hepi.ac.uk](https://www.hepi.ac.uk/reports/student-generative-ai-survey-2026/), 2026
- [Deployment] Nigeria · Student — 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. [openknowledge.worldbank.org](https://openknowledge.worldbank.org/entities/publication/15e1ff08-15ae-4f7a-b2a8-d146e6c113ee/full), 
- [Evidence of adoption] Global — 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. [oecd.org](https://www.oecd.org/content/dam/oecd/en/publications/reports/2025/04/bridging-the-ai-skills-gap_b43c7c4a/66d0702e-en.pdf), 2024
- [Evidence of adoption] Global — 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. [reports.weforum.org](https://reports.weforum.org/docs/WEF_Future_of_Jobs_Report_2025.pdf), 2025
- [Evidence of adoption] Singapore — 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. [isomer-user-content.by.gov.sg](https://isomer-user-content.by.gov.sg/41/719a22ec-2e40-49f8-8d3a-1469f90efdd3/SSG%20Annual%20Report%20FY24-25.pdf), 2024
- [Deployment] India · Learner / trainee — 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. [pib.gov.in](https://www.pib.gov.in/PressReleasePage.aspx?PRID=2146573&reg=48&lang=2), 2024
- [Evidence of impact] Turkey — 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. [pnas.org](https://www.pnas.org/doi/10.1073/pnas.2422633122), 2026
- [Evidence of adoption] United States — 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. [news.gallup.com](https://news.gallup.com/poll/710534/teachers-receive-no-formal-guidance.aspx), 2026
- [Evidence of impact] Global — 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. [link.springer.com](https://link.springer.com/article/10.1007/s40979-026-00226-w), 2026
- [Evidence of adoption] Global — 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. [oecd.org](https://www.oecd.org/en/publications/oecd-digital-education-outlook-2026_062a7394-en.html), 2026
- [Evidence of adoption] Global — 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. [oecd.org](https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/06/developing-vocational-education-and-training-with-artificial-intelligence_b7efe1ae/e9f76b4e-en.pdf), 2025
- [Evidence of adoption] Global — 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. [oecd.org](https://www.oecd.org/content/dam/oecd/en/publications/support-materials/2026/06/developing-vocational-education-and-training-with-artificial-intelligence_b7efe1ae/Ten-country-case-studies-on-VET-development-and-AI-use.pdf), 2025
- [Regulation] United Kingdom — 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. [skillsengland.education.gov.uk](https://skillsengland.education.gov.uk/apprenticeships/st1512-v2-1), 2025
- [Regulation] United Kingdom — 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. [skillsengland.education.gov.uk](https://skillsengland.education.gov.uk/apprenticeship-units/AU0009), 2026
- [Regulation] India — 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. [pib.gov.in](https://www.pib.gov.in/PressReleasePage.aspx?PRID=2287346&reg=1&lang=1), 2026
- [Regulation] United States — 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. [dol.gov](https://www.dol.gov/agencies/eta/advisories/ten-07-25), 2026
- [Deployment] United States · Apprentice — 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. [apprenticeship.gov](https://www.apprenticeship.gov/AI-in-registered-apprenticeships), 2026
- [Regulation] Singapore — 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. [swda.gov.sg](https://www.swda.gov.sg/home/newsroom/appointment-of-inaugural-board-for-skills-and-workforce-development-agency), 2026
- [Deployment] Singapore · Learner — 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. [mom.gov.sg](https://www.mom.gov.sg/newsroom/press-releases/2026/0303-partnering-businesses-and-workers-to-thrive-in-a-changed-world), 2026
- [Deployment] Estonia · Vocational student — 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. [hm.ee](https://www.hm.ee/en/news/estonia-announces-groundbreaking-national-initiative-ai-leap-programme-bring-ai-tools-all), 2026
- [Evidence of adoption] European Union — 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. [cedefop.europa.eu](https://www.cedefop.europa.eu/files/6229_en.pdf), 2025
- [Evidence of adoption] European Union — 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. [cedefop.europa.eu](https://www.cedefop.europa.eu/files/9209_en.pdf), 2026
- [Evidence of impact] Singapore — 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. [arxiv.org](https://arxiv.org/abs/2512.12306), 2025
- [Evidence of impact] United Kingdom — 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. [gov.uk](https://www.gov.uk/government/publications/ai-in-schools-and-further-education-findings-from-early-adopters/the-biggest-risk-is-doing-nothing-insights-from-early-adopters-of-artificial-intelligence-in-schools-and-further-education-colleges), 2026
- [Evidence of adoption] United Kingdom — 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. [assets.publishing.service.gov.uk](https://assets.publishing.service.gov.uk/media/683729a7dc6ebc5eca0cbb29/Evaluation_of_skills_bootcamps_-_2022_to_2023_-_completions_and_outcomes_report.pdf), 2023
- [Evidence of adoption] India — 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. [cag.gov.in](https://cag.gov.in/uploads/download_audit_report/2025/Report-No.-20-of-2025_PA-PMKVY_English-PDF-A-06943abec463479.68516873.pdf), 2022
- [Evidence of adoption] European Union — 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%. [cedefop.europa.eu](https://www.cedefop.europa.eu/en/data-insights/adult-education-and-training-statistics-overview), 2022
- [Evidence of adoption] Global — 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. [oecd.org](https://www.oecd.org/content/dam/oecd/en/publications/reports/2026/07/skills-in-the-ai-age_e8d8c1e6/972bd15e-en.pdf), 2024
- [Evidence of adoption] United Kingdom — 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. [multiverse.io](https://www.multiverse.io/blog/multiverse-raises-70-million-europes-ai-adoption-platform), 2026
- [Evidence of adoption] India — 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. [indiaai.gov.in](https://indiaai.gov.in/news/cabinet-approves-india-ai-mission-at-an-outlay-of-rs-10-372-crore), 2026
- [Evidence of adoption] Singapore — 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. [content.mycareersfuture.gov.sg](https://www.content.mycareersfuture.gov.sg/budget-2026-singaporean-workers-employers), 2026
- [Deployment] Haiti · examinations officer — 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. [haitilibre.com](https://www.haitilibre.com/article-48137-haiti-technologie-l-ia-dans-le-controle-qualite-avant-la-validation-finale-des-resultats-d-examens-d-etat.html), 2026-08-04
- [Deployment] Trinidad and Tobago · secondary school teacher — 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. [ttt.live](https://www.ttt.live/education-ministry-launches-new-digital-learning-platform-to-support-secondary-school-students/), 2026-09-02
- [Deployment] United States · school security officer — 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. [globenewswire.com](https://www.globenewswire.com/news-release/2026/08/25/3350361/0/en/wamego-usd-320-deploys-zeroeyes-ai-gun-detection-and-intelligent-situational-awareness-platform.html), 2026-08-25
- [Deployment] United States · university lecturer — 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. [patch.com](https://patch.com/illinois/beverly-mtgreenwood/saint-xavier-university-debuts-bold-ai-learning-platform-campus-wide), 2026-08-18
- [Regulation] United States — 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. [cbsnews.com](https://www.cbsnews.com/miami/news/florida-state-board-education-ai-school-rule-vote/), 2026-08-26
- [Deployment] Singapore · Teacher — 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. [redhot.sg](https://www.redhot.sg/new-ai-tool-helps-teachers-mark-geography-essays-english-history-and-social-studies-being-piloted-713830.html), 2026-09-05
- [Deployment] Canada · University lecturer — 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. [news.ontariotechu.ca](https://news.ontariotechu.ca/archives/2026/02/ontario-tech-launches-human-centred-ai-learning-agent-pilot.php), 2026-02
- [Deployment] Morocco · Exam invigilator — 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. [thevoice.ma](https://thevoice.ma/%D8%A7%D9%84%D8%A8%D9%83%D8%A7%D9%84%D9%88%D8%B1%D9%8A%D8%A7-2026-%D8%B0%D9%83%D8%A7%D8%A1-%D8%A7%D8%B5%D8%B7%D9%86%D8%A7%D8%B9%D9%8A-%D9%84%D8%B1%D8%B5%D8%AF-%D8%A7%D9%84%D8%BA%D8%B4-%D9%88%D8%AE/), 2026-06
- [Regulation] Brazil — 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. [olhardigital.com.br](https://olhardigital.com.br/2026/09/01/inteligencia-artificial/ia-nao-podera-mais-corrigir-provas-e-redacoes-nas-escolas-veja-o-que-muda), 2026-09-01
- [Evidence of adoption] Dominican Republic — 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. [diariolibre.com](https://www.diariolibre.com/actualidad/educacion/2026/09/08/uso-de-dispositivos-digitales-en-educacion-y-su-impacto/3652813), 2026-09-08
- [Evidence of impact] El Salvador — 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. [infobae.com](https://www.infobae.com/el-salvador/2026/09/09/programa-piloto-con-ia-y-tecnologia-lleva-a-171-escuelas-salvadorenas-a-niveles-de-paises-desarrollados-en-pisa-dice-el-gobierno/), 2026-09-09
- [Regulation] Norway — 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. [barnevakten.no](https://www.barnevakten.no/ki-skal-bort-fra-barneskolene/), 2026-06-19
- [Deployment] France · Trainee bank and wealth-management adviser — 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. [esteval.fr](https://www.esteval.fr/article.47586.initiatives-sopra-steria-et-gwenlake-dotent-l-ecole-superieure-de-la-banque-d-un-entraineur-ia), 2026-08-31
- [Deployment] Germany · Teacher — 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. [fwu.de](https://fwu.de/presse/telli-in-weiterem-bundesland-im-einsatz-niedersachsen-fuehrt-ki-chatbot-an-schulen-ein/), 2026-02-16
- [Deployment] Germany · Teacher — 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. [news4teachers.de](https://www.news4teachers.de/2026/04/bildungsminister-sieht-historischen-schritt-ki-chatbot-fuer-den-unterricht-freigeschaltet/), 2026-04-03
- [Deployment] Spain · Teacher — 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. [telemadrid.es](https://www.telemadrid.es/noticias/madrid/La-IA-llega-a-EducaMadrid-para-que-los-profesores-creen-contenidos-personalizados-para-los-alumnos-2-2850634970--20260107120100.html), 2026-01-08
- [Evidence of adoption] Mexico — 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. [merca20.com](https://www.merca20.com/prueba-pisa-mide-el-uso-de-ia-entre-estudiantes-mexicanos-47-ya-usa-chatbots-al-menos-una-vez-por-semana/), 2026-09-09
- [Deployment] Ethiopia · Secondary school teacher — 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. [addisfortune.news](https://addisfortune.news/ministry-of-education-huawei-introduce-offline-ai-support-system), 2026-05
- [Deployment] Nigeria · Teacher — 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. [ogtv.com.ng](https://ogtv.com.ng/2026/01/22/ogun-launches-first-ai-powered-teaching-platform/), 2026-01-22
- [Deployment] Nigeria · Primary school teacher — 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. [pulse.ng](https://www.pulse.ng/story/jigawa-deploys-ai-system-that-teaches-public-school-pupils-from-home-when-they-miss-school-2026061415040751227), 2026-06-14
- [Deployment] Nigeria · Primary school teacher — 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. [thisdaylive.com](https://www.thisdaylive.com/2026/03/16/aibom-launches-ai-driven-basic-education-reform-to-boost-foundational-learning/), 2026-03-16
- [Deployment] Mauritius · Teacher — 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. [wearetech.africa](https://www.wearetech.africa/en/fils-uk/news/tech/mauritius-rolls-out-generative-ai-tool-in-public-schools), 2026-02-02
- [Regulation] Seychelles — 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. [statehouse.gov.sc](https://statehouse.gov.sc/news/7158/president-herminies-vision-for-artificial-intelligence-takes-root-in-seychelles-classrooms), 2026-07-31
- [Deployment] Panama · Judicial training instructor — 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. [organojudicial.gob.pa](https://www.organojudicial.gob.pa/isjup/noticias/el-isjup-impulsa-la-innovacion-judicial-con-el-lanzamiento-de-su-asistente-virtual-basado-en-inteligencia-artificial), 2026-03-20
- [Deployment] Trinidad and Tobago · Secondary school teacher — 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. [guardian.co.tt](https://www.guardian.co.tt/news/dowlath-launches-ai-textbooks-for-form-1-students-to-help-cut-costs-for-parents-6.2.2567856.18fe38a59e), 2026-04-22
- [Evidence of adoption] Uruguay — 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. [anep.edu.uy](https://www.anep.edu.uy/en-uruguay-tres-cada-cuatro-docentes-utilizan-ia-para-generar-propuestas-educativas-pedagogicas), 2026-05-12
- [Regulation] Uruguay — 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. [ladiaria.com.uy](https://ladiaria.com.uy/educacion/articulo/2026/5/udelar-aprobo-principios-rectores-para-el-uso-de-inteligencia-artificial-en-la-institucion/), 2026-05-27
- [Deployment] Uruguay · School teacher — 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. [gub.uy](https://www.gub.uy/presidencia/comunicacion/noticias/inteligencia-artificial-docentes-estudiantes-ceibal-anep-nuevas-2026), 2026-06-16
- [Deployment] Argentina · University lecturer — 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. [lanacion.com.ar](https://www.lanacion.com.ar/sociedad/herramientas-digitales-una-facultad-de-la-uba-regulo-el-uso-de-ia-en-las-aulas-y-lanzo-una-nid27062026/), 2026-06-27
- [Deployment] Brazil · School teacher — 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. [diarionline.com.br](https://diarionline.com.br/?s=noticia&id=157303), 2026-06-01
- [Deployment] Kuwait · Secondary school teacher — 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. [arabtimesonline.com](https://www.arabtimesonline.com/news/kuwaits-education-ministry-rolls-out-ai-tools-program-with-google/), 2026-02-11
- [Regulation] United Arab Emirates — 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. [emaratalyoum.com](https://www.emaratalyoum.com/local-section/education/2026-02-21-1.2017892), 2026-02-21
- [Deployment] Palestine · Medical school examiner — 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. [nn.najah.edu](https://nn.najah.edu/news/Alnajah-news/2026/04/10/554144/), 2026-04-10
- [Deployment] Egypt · Teacher — 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. [egyptianstreets.com](https://egyptianstreets.com/2026/05/20/egypt-launches-ai-based-adhd-assessment-tool-for-children/), 2026-05-20
- [Regulation] Egypt — 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. [unesco.org](https://www.unesco.org/en/articles/national-artificial-intelligence-competency-framework-teachers-launched-egypt), 2026-07-21
- [Deployment] Algeria · University guidance counsellor — 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. [aps.dz](https://www.aps.dz/fr/algerie/education-et-technologie/mrdpiw05-mise-en-ligne-de-la-circulaire-ministerielle-numerique-et-de-l-assistant-ia-d-orientation), 2026-07-09
- [Deployment] Morocco · Examination marker — 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. [al3omk.com](https://al3omk.com/1160047.html), 2026-05-11
- [Deployment] Turkey · Foreign-language teacher — 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. [yegitek.meb.gov.tr](https://yegitek.meb.gov.tr/www/yapay-zeka-destekli-yabanci-dil-platformu-quotdilimquot-220-binden-fazla-kullaniciya-ulasti/icerik/3999), 2026-07-19
- [Regulation] Turkey — 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. [ttkb.meb.gov.tr](https://ttkb.meb.gov.tr/www/egitimde-yapay-zeka-uygulamalari-icin-etik-beyan-sureci-basladi-yazek-yayinda/icerik/864), 2026-02-02
- [Deployment] Israel · English teacher — 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. [ynet.co.il](https://www.ynet.co.il/news/article/h1ezq3iwzg), 2026-06-10
- [Regulation] United States — 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. [nyc.gov](https://www.nyc.gov/mayors-office/news/2026/09/mayor-mamdani-and-chancellor-samuels-put-students-first-with-nat), 2026-09-02
- [Regulation] United States — 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. [k12dive.com](https://www.k12dive.com/news/lausd-restricts-all-students-from-using-ai-tools/829612/), 2026-09-03
- [Deployment] United States · university lecturer — 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. [kiowacountypress.net](https://kiowacountypress.net/content/iowa-university-launches-generative-ai-pilot-program-tools-access-community-conversations), 2026-09-08
- [Deployment] Canada · Secondary school teacher — 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. [martlet.ca](https://martlet.ca/does-a-thirteen-year-old-really-need-an-ai-chatbot-account-in-school/), 2026-06
- [Deployment] United States · University lecturer — 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. [uit.stanford.edu](https://uit.stanford.edu/news/new-ai-tools-stanford-arrive-june-30), 2026-06-24
- [Evidence of adoption] Lebanon — 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. [ul.edu.lb](https://www.ul.edu.lb/en/node/6453), 2026-07-28
- [Regulation] Albania — 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. [timoni.al](https://timoni.al/chatgpt-ligjerohet-ne-shkolla-udhezuesi-i-ri-i-ministrise-se-arsimit-percakton-perdorimin-ne-ushtrime-dhe-vleresim-nxenesish/), 2026-09-07
- [Regulation] Benin — 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. [siauac.sigan-uac.bj](https://siauac.sigan-uac.bj/), 2026-07-16
- [Deployment] Armenia · Teacher — 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. [armeniaeducationinitiative.org](https://www.armeniaeducationinitiative.org/en/seroond), 2026-09-01
- [Deployment] Armenia · Teacher — 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. [armenpress.am](https://armenpress.am/en/article/1253791), 2026-06-23
- [Regulation] Belarus — 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. [pravo.by](https://pravo.by/novosti/obshchestvenno-politicheskie-i-v-oblasti-prava/2026/february/92336/), 2026-02-17
- [Evidence of adoption] Syria — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC13186531/), 2026-05-19
- [Deployment] Azerbaijan · Teacher — 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. [edu.gov.az](https://edu.gov.az/en/announcements/23004-1), 2026-04-29
- [Deployment] Barbados · Secondary school student — 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. [barbadostoday.bb](https://barbadostoday.bb/2026/05/15/students-gain-investing-experience-through-stock-market-ai-programme/), 2026-05-15
- [Regulation] Barbados — 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. [barbadostoday.bb](https://barbadostoday.bb/2026/07/17/cxc-to-replace-sbas-with-invigilated-paper-amid-ai-concerns/), 2026-07-17
- [Evidence of adoption] Croatia — 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'. [dalmacijadanas.hr](https://www.dalmacijadanas.hr/umjetna-inteligencija-ulazi-u-skole-evo-o-cemu-ce-daci-uciti/), 2026-09-01
- [Evidence of adoption] Yemen — 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. [moe-ye.net](https://moe-ye.net/36617), 2026-07-05
- [Deployment] Cameroon · Secondary school teacher — 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. [fr.journalducameroun.com](https://fr.journalducameroun.com/cameroun-lia-le-nouveau-laboratoire-des-lycees-camerounais/), 2026-08-19
- [Evidence of adoption] Czechia — 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. [isic.cz](https://www.isic.cz/clanek/ai-pouziva-pro-skolni-ucely-devet-z-deseti-studentu-ale-pravidla-ma-jen-petina-skol), 2026-06-25
- [Deployment] Bermuda · Teacher — 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. [bernews.com](https://bernews.com/2026/05/chatmore-introduces-ai-learning-platform/), 2026-05-08
- [Deployment] Estonia · Upper-secondary student — 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. [eki.ee](https://eki.ee/uudis/ti-huppe-tehisaru-opirakenduse-nimekonkursi-voitis-ulekaalukalt-iti/), 2026-05-12
- [Evidence of adoption] Cambodia — 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. [mptc.gov.kh](https://mptc.gov.kh/en/2026/07/members-of-the-club-of-cambodian-journalists-received-a-capacity-building-training-on-digital-skills-and-artificial-intelligence-to-boost-productivity-and-work-efficiency-in-cambodias-digital-journa/), 2026-07-28
- [Deployment] Guernsey · Teacher — 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. [guernseypress.com](https://guernseypress.com/news/2026/06/25/elizabeth-college-trialling-new-ai-learning-tutor-tool), 2026-06-25
- [Evidence of adoption] Cook Islands — 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. [cookislandsnews.com](https://www.cookislandsnews.com/internal/national/technology/young-cook-islander-empowers-workforce-through-kia-orana-ai-programme/), 2026-02-06
- [Deployment] DR Congo · Exam marker — 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. [edu-nc.gouv.cd](https://edu-nc.gouv.cd/actualites/examen-d-etat-2026-les-premiers-resultats-attendus-des-ce-weekend-annonce-l-inspecteur-general), 2026-06-26
- [Evidence of adoption] Cyprus — 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. [cyprus-mail.com](https://cyprus-mail.com/2026/02/16/cyprus-youth-among-top-ai-users-as-generative-tools-surge-in-2025), 2026-02-16
- [Evidence of adoption] Equatorial Guinea — 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. [ahoraeg.com](https://ahoraeg.com/economia/empresa/2026/09/11/bange-capacita-a-su-personal-en-inteligencia-artificial-y-fintech-para-modernizar-su-actividad-bancaria/), 2026-09-11
- [Deployment] Hungary · Teacher — 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. [telex.hu](https://telex.hu/belfold/2026/01/28/mesterseges-intelligencia-intezmenyi-oktatas-szakkepzes-iskolak-tanarok-mi-psz), 2026-01-28
- [Deployment] Kazakhstan · Primary school teacher — 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. [timesca.com](https://timesca.com/kazakhstan-ai-education-schools-university/), 2026-08-18
- [Deployment] Philippines · University lecturer — 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. [gadgetsmagazine.com.ph](https://gadgetsmagazine.com.ph/lifestyle/education/ust-gemini-for-education), 2026-09-14
- [Deployment] Gambia · secondary school students, teachers — 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. [thepoint.gm](https://thepoint.gm/africa/gambia/national-news/cube-launches-pilot-ai-learning-tool-in-gambian-secondary-schools), 2026-05-08
- [Deployment] Guinea · teacher — 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. [guinee7.com](https://www.guinee7.com/2026/05/08/education-une-intelligence-artificielle-bientot-au-service-des-candidats-au-bac/), 2026-05-08
- [Deployment] Madagascar · Primary school teachers — 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. [fhi360.org](https://www.fhi360.org/articles/new-ai-chatbot-aims-to-boost-teacher-pedagogical-support/), 2026-05-11
- [Deployment] Belarus · Computer science teacher — 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. [edu.gov.by](https://edu.gov.by/news/s-1-sentyabrya-uchashchiesya-belarusi-nachnut-izuchat-iskusstvennyy-intellekt-utverzhdena-obnovlyenn/), 29 Jul 2026
- [Deployment] Czechia · University lecturer — 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. [muni.cz](https://www.muni.cz/pro-media/tiskove-zpravy/masarykova-univerzita-jako-prvni-v-cesku-zavadi-ai-do-sveho-informacniho-systemu), 10 Sep 2026
- [Evidence of adoption] Guernsey — 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. [channeleye.media](https://channeleye.media/guernsey-goes-all-in-ai-sprint-passes-1000-bookings-in-under-a-week/), 9 Sep 2026
- [Evidence of adoption] Greece — 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. [alfavita.gr](https://www.alfavita.gr/koinonia/559793_bgainoyn-ayto-mina-oi-protoi-stin-ellada-pistopoiimenoi-dimosioi-ypalliloi-stin), 10 Sep 2026
- [Deployment] Ukraine · teacher — 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. [unn.ua](https://unn.ua/en/news/in-ukraine-ai-will-evaluate-students-assignments-through-the-mriya-application), 2026-01-29
- [Deployment] North Korea · Teachers — 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. [nkeconomy.com](http://www.nkeconomy.com/news/articleView.html?idxno=15860), 2026-01-22
- [Evidence of adoption] Papua New Guinea — 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. [thenational.com.pg](https://www.thenational.com.pg/ai-is-in-the-classroom-insights-from-a-survey-of-upng-students/), 2026-06-15
- [Deployment] Senegal · Teacher, student — 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. [osiris.sn](https://www.osiris.sn/allo-gor-gui-l-ia-edu-ca-tive-qui-rap-proche-le-savoir-des-eleves.html), 2026-07-31
- [Deployment] Sierra Leone · lecturers — 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. [techreviewafrica.com](https://techreviewafrica.com/news/5093/sierra-leone-pilots-ai-driven-online-examination-system-to-enhance-higher-education), 2026-04-23
- [Regulation] Somalia — 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. [frontieronline.co.ke](https://frontieronline.co.ke/simad-university-adopts-first-formal-ai-policy-as-somalia-pushes-for-responsible-artificial-intelligence/), 2026-08-16
- [Evidence of adoption] Somalia — 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. [iafrica.com](https://iafrica.com/somalias-bile-initiative-trains-160-final-year-ict-students-with-ai-among-four-tracks-and-no-foreign-backer/), 2026-08-24
- [Evidence of adoption] Somalia — 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. [mu.edu.so](https://mu.edu.so/mogadishu-university-launches-cyber-security-and-artificial-intelligence-degree-program/), 2026-08-13
- [Deployment] South Sudan · teacher, student — 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. [radioyei.org](https://radioyei.org/news/technology/data-free-junub-academy-brings-ai-supported-education-to-south-sudan/), 2026-06-29
- [Evidence of adoption] Guadeloupe — 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. [fr.linkedin.com](https://fr.linkedin.com/posts/universit-des-antilles_luniversit%C3%A9-des-antilles-lance-son-dipl%C3%B4me-activity-7430612555481395201-wGkx), 2026-02-20
- [Deployment] Saint Lucia · admissions officers — 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. [salcc.edu.lc](http://www.salcc.edu.lc/news/salcc-launches-ai-chatbot.php), 2026-02-16
- [Regulation] Virgin Islands, U.S. — 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. [uvi.edu](https://www.uvi.edu/news/2026/26_094_issue.html), 2026-09-04
- [Deployment] Canada · teacher — 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. [ici.radio-canada.ca](https://ici.radio-canada.ca/nouvelle/2168255/emilia-ia-correction-professeur-ecole), 2025-05-29
- [Deployment] Canada · teacher / school counsellor — 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. [engaged-learning.com](https://engaged-learning.com/ai-tool-to-prevent-dropout-quebec-school-service-centres/), 2025-05-08
- [Deployment] Canada · teacher / student — 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. [journalexpress.ca](https://www.journalexpress.ca/actualite/quatre-ia-generatives-integrees-au-recueil-doutils-des-eleves-et-du-personnel/), 2025-09-04

## Retail & Consumer

- [Deployment] United States · Customer-service agent — 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. [retaildive.com](https://retaildive.com/news/best-buy-ai-shopping-assistant/829049/), 31 Aug 2026
- [Deployment] United States · Supply-chain planner — 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. [supplychaindive.com](https://www.supplychaindive.com/news/dollar-general-deploys-ai-across-distribution-centers-stores/828966/), 27 Aug 2026
- [Deployment] United Kingdom · Supply-chain / agri — 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. [grocerygazette.co.uk](https://www.grocerygazette.co.uk/2026/08/27/ms-rolls-out-ai-powered-technology-across-dairy-farms/), 27 Aug 2026
- [Deployment] United Kingdom · Loss prevention — 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. [retailgazette.co.uk](https://www.retailgazette.co.uk/blog/2026/08/morrisons-rolls-out-ai-powered-self-checkouts-to-200-stores/), 5 Aug 2026
- [Deployment] United Kingdom · Marketer — 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. [retailtechinnovationhub.com](https://retailtechinnovationhub.com/home/2026/8/13/allsaints-goes-live-with-new-customer-engagement-platform-powered-by-klaviyo-technology), 13 Aug 2026
- [Deployment] United States · Store operations — 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. [retailtechinnovationhub.com](https://retailtechinnovationhub.com/home/2026/8/9/milestone-for-simbes-tally-solution-as-500th-robot-goes-live-across-krogers-store-network), 9 Aug 2026
- [Deployment] United Kingdom · Loss prevention — 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. [grocerygazette.co.uk](https://www.grocerygazette.co.uk/2026/08/04/asda-rolls-out-loss-prevention-tool-across-full-estate-following-100k-savings/), 4 Aug 2026
- [Deployment] South Korea · Merchandiser — 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. [en.sedaily.com](https://en.sedaily.com/finance/2026/08/26/lotte-department-store-rolls-out-ai-to-scout-new-brands), 2026-08-26
- [Deployment] Belgium · Customer-facing / retail ops — 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. [retaildetail.eu](https://www.retaildetail.eu/news/food/carrefour-brings-ai-shopping-assistant-to-belgium/), 2026-08-25
- [Deployment] Netherlands · Customer service — 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. [retailtechinnovationhub.com](https://retailtechinnovationhub.com/home/2026/8/11/just-eat-takeawaycom-rolls-out-ai-voice-assistant-across-europe-building-on-uk-and-germany-launches), 2026-08-11
- [Deployment] United Kingdom · Loss prevention — 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. [retailtechinnovationhub.com](https://retailtechinnovationhub.com/home/2026/8/30/sainsburys-says-sorry-and-walmart-admits-defeat-the-stand-out-retail-technology-plays-from-august), 2026-08-30
- [Deployment] Canada · Staff / merchandiser — 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. [retail-insider.com](https://retail-insider.com/retail-insider/2026/08/canadian-tire-prepares-for-ai-shopping-as-retail-strategy-evolves/), 2026-08-20
- [Deployment] Poland · Marketplace seller — 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. [about.allegro.eu](https://about.allegro.eu/wp-content/uploads/2026/05/Komunikat-prasowy-Q1-2026.pdf), 2026-05
- [Deployment] Canada · Store manager — 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. [loblaw.ca](https://www.loblaw.ca/en/loblaw-advances-ai-in-canadian-retail-with-first-of-its-kind-shopping-app-in-chatgpt/), 2026-02-12
- [Deployment] Poland · Call-centre consultant — 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. [itreseller.pl](https://itreseller.pl/allegro-i-elevenlabs-uruchamiaja-glosowego-agenta-ai-na-infolinii-allegro-delivery/), 2026-09-07
- [Deployment] United Arab Emirates · Hotel receptionist — 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. [hotelmanagement-network.com](https://www.hotelmanagement-network.com/news/h2o-check-in-rotana-hotels/), 2026-08-20
- [Deployment] Australia · Supermarket team member — 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. [itnews.com.au](https://www.itnews.com.au/news/woolworths-gives-agentic-powered-olive-chatbot-to-its-200000-staff-625339), 2026-04-28
- [Deployment] Guatemala · Retail sales adviser — 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. [emisorasunidas.com](https://emisorasunidas.com/empresas/2026/04/14/cemaco-presenta-a-maco-su-innovador-asistente-virtual-de-inteligencia-artificial), 2026-04-14
- [Deployment] United States · retail store associate — 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. [corporate.homedepot.com](https://corporate.homedepot.com/news/company/home-depot-expands-magic-apron-deliver-personalized-localized-store-guidance), 2026-08-27
- [Deployment] Anguilla · Tourism information officer — 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. [theanguillian.com](https://theanguillian.com/2026/05/anguilla-tourist-board-unveils-bold-summer-campaign-and-innovative-xcape-app/), 2026-05-27
- [Deployment] Czechia · Customer service agent — 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. [zboziaprodej.cz](https://www.zboziaprodej.cz/2026/05/18/rohlik-group-integruje-platby-primo-do-ai-asistenta-nakoupit-lze-i-hlasem/), 2026-05-18
- [Deployment] Czechia · E-commerce merchandiser — 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. [ceskenoviny.cz](https://www.ceskenoviny.cz/tiskove/zpravy/rohlik-prinasi-jednu-z-prvnich-nakupnich-aplikaci-pro-chatgpt-v-evrope/2851011), 2026-07-14
- [Deployment] Czechia · Data analyst — 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. [mam.cz](https://www.mam.cz/novinky/kreativita-a-kampane/2026-06/umela-inteligence-usetri-az-20-personalnich-nakladu-e-shop-mironet-meni-komunikaci-a-pripravuje-revolucni-vyhledavani/), 2026-06-18
- [Deployment] Brunei · Tour operator — 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. [thescoop.co](https://thescoop.co/2026/06/30/brunei-launches-nakhoda-to-steer-tourism-industry-into-the-digital-age/), 2026-06-30
- [Deployment] Slovakia · shop assistant — 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. [tovarapredaj.sk](https://www.tovarapredaj.sk/2026/03/11/metro-slovakia-predstavilo-prvu-hybridnu-autonomnu-predajnu-v-sieti-moj-obchod/), 2026-03-11
- [Deployment] Belgium · Checkout assistant — 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. [retaildetail.be](https://www.retaildetail.be/nl/news/food/carrefour-belgie-test-slimme-winkelkar-van-instacart/), 10 Sep 2026
- [Deployment] Czechia · E-commerce manager — 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. [lupa.cz](https://www.lupa.cz/clanky/seznam-na-zbozi-cz-spousti-nakupovani-pomoci-ai-agentu/), 1 Sep 2026
- [Deployment] Switzerland · retail sales associate — 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. [srf.ch](https://www.srf.ch/news/wirtschaft/ki-im-laden-swisscom-nimmt-kundengespraeche-auf-das-muessen-sie-wissen), 2026-07-10
- [Deployment] Åland Islands · tourism entrepreneur — 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. [alandsradio.ax](https://alandsradio.ax/nyheter/hogskolan-satsar-20-000-pa-ai-verktyg-for-turistforetagare), 2026-09-01
- [Deployment] South Sudan · food and beverage staff — 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. [itweb.africa](https://itweb.africa/article/ayen-becomes-first-robot-waitress-in-south-sudan/RgeVDvPRgwlMKJN3), 2026-01-15

## Media, Telecom & Entertainment

- [Deployment] United States · Customer-service agent — 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. [rcrwireless.com](https://rcrwireless.com/20260826/ai/verizon-google-cloud-ai-partnership), 26 Aug 2026
- [Deployment] United Kingdom · Advertising / creative — 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. [storyboard18.com](https://www.storyboard18.com/how-it-works/wpp-omnicom-get-first-mover-edge-on-chatgpt-ads-but-self-serve-model-puts-agency-role-under-scrutiny-ws-o-108931.htm), 27 Aug 2026
- [Deployment] United States · Journalist — 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. [aiagentstore.ai](https://aiagentstore.ai/ai-agent-news/daily/2026-08-29), 29 Aug 2026
- [Deployment] Spain · Customer service — 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. [rcrwireless.com](https://rcrwireless.com/20260812/ai/telefonica-ai-voice-services), 2026-08-12
- [Deployment] Germany · Broadcast production — 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. [broadcast-solutions.de](https://www.broadcast-solutions.de/en/news), 2026-09-01
- [Deployment] Canada · Network / customer — 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. [mobilesyrup.com](https://mobilesyrup.com/2026/07/28/bell-ai-powered-anti-spoofed-call-system/), 2026-07-28
- [Evidence of impact] Global — 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. [reutersinstitute.politics.ox.ac.uk](https://reutersinstitute.politics.ox.ac.uk/journalism-media-and-technology-trends-and-predictions-2026), 2026
- [Evidence of adoption] Global — Researchers' AI tool use rose to 84% in 2025 from 57% in 2024. Survey: 2,430 researchers, August 2025. [newsroom.wiley.com](https://newsroom.wiley.com/press-releases/press-release-details/2025/AI-Adoption-Jumps-to-84-Among-Researchers-as-Expectations-Undergo-Significant-Reality-Check/default.aspx), 2025
- [Evidence of impact] Global — 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. [newsroom.wiley.com](https://newsroom.wiley.com/press-releases/press-release-details/2025/AI-Adoption-Jumps-to-84-Among-Researchers-as-Expectations-Undergo-Significant-Reality-Check/default.aspx), 2025
- [Deployment] United Kingdom · Journalist — 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. [wan-ifra.org](https://wan-ifra.org/2024/10/how-uks-reach-is-using-ai-to-help-produce-more-content-faster/), 
- [Deployment] South Korea · Stadium operations staff — 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. [en.sedaily.com](https://en.sedaily.com/finance/2026/09/06/kt-deploys-ai-at-baseball-park-to-curb-scalping-translate), 2026-09-06
- [Deployment] Nigeria · Journalist — 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. [ipcng.org](https://www.ipcng.org/osundecides2026-ipc-unveils-ai-powered-waflo-to-strengthen-journalists-safety/), 2026-08-14
- [Evidence of impact] Jamaica — 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. [jamaica-gleaner.com](http://jamaica-gleaner.com/article/business/20260821/digicel-set-roll-out-ai-agent-ruby-across-caribbean-after-jamaica-debut), 2026-08-21
- [Deployment] Colombia · Call-centre agent — 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. [crcom.gov.co](https://www.crcom.gov.co/es/noticias/comunicado-prensa/inicia-fase-experimentacion-contigo-digital), 2026-06-18
- [Regulation] Mexico — 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. [elimparcial.com](https://www.elimparcial.com/mexico/2026/05/15/ya-es-oficial-la-nueva-ley-protege-a-locutores-y-actores-de-doblaje-contra-la-clonacion-digital-con-multas-de-hasta-40-mil-dias-de-salario-por-usar-inteligencia-artificial-para-suplantar-su-labor-profesional-sin-consentimiento/), 2026-05-15
- [Deployment] Mexico · Dubbing actor — 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. [xataka.com.mx](https://www.xataka.com.mx/robotica-e-ia/televisa-encontro-clave-para-llevar-sus-telenovelas-a-paises-ahora-dobla-ia), 2026-07-29
- [Deployment] Qatar · News translator — 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. [qna.org.qa](https://qna.org.qa/en/News-Area/News/2026-5/10/qna-launches-news-services-in-three-new-languages-using-ai), 2026-05-10
- [Deployment] Lebanon · Journalist — 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. [arabnews.com](https://www.arabnews.com/node/2651419/media), 2026-07-18
- [Deployment] Lebanon · News presenter — 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. [safirpress.net](https://www.safirpress.net/2026/07/28/%D8%A3%D9%88%D9%91%D9%84-%D8%AA%D9%81%D8%A7%D8%B9%D9%84-%D8%A8%D9%8A%D9%86-%D9%85%D8%B0%D9%8A%D8%B9%D8%A9-%D9%88%D9%85%D9%8F%D8%B1%D8%A7%D8%B3%D9%84-%D8%B1%D9%82%D9%85%D9%8A%D9%91%D9%8A%D9%92%D9%86/), 2026-07-28
- [Deployment] Andorra · Filmmaker — 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. [alto.ad](https://www.alto.ad/culture/andorra-referents-exhibition-lita-cabellut--260331), 2026-04-02
- [Regulation] Benin — 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. [lanouvelletribune.info](https://lanouvelletribune.info/2026/03/presidentielle-au-benin-ce-quil-faut-savoir-des-regles-fixees-pour-la-campagne-mediatique/), 2026-03-13
- [Deployment] Armenia · Animator — 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. [armenianweekly.com](https://armenianweekly.com/2026/06/25/little-mher-brings-armenian-epic-into-ai-age/), 2026-06-25
- [Deployment] Belarus · Photo editor — 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. [belta.by](https://belta.by/society/view/nejroset-instrument-a-ne-mozg-prinimajuschij-reshenija-v-globalnom-smysle-proekty-belta-sozdannye-s-784286-2026/), 2026-06-03
- [Evidence of adoption] Yemen — 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%. [althawrah.ye](https://althawrah.ye/archives/1234609), 2026-09-09
- [Deployment] Cabo Verde · Voter — 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. [inforpress.cv](https://www.inforpress.cv/pt/projecto-kriol-think-lanca-assistente-de-ia-para-comparar-programas-eleitorais), 2026-05-08
- [Deployment] Central African Republic · Fact-checking journalist — 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. [radiondekeluka.org](https://www.radiondekeluka.org/98693-attention-la-coree-du-nord-nest-pas-entree-dans-la-guerre-en-iran), 2026-04-02
- [Deployment] Central African Republic · Fact-checking journalist — 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. [centrafriquecheck.org](https://centrafriquecheck.org/cette-photo-est-generee-par-lintelligence-artificielle/), 2026-03
- [Deployment] Faroe Islands · Journalist — 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. [kvf.fo](https://kvf.fo/greinar/2026/04/23/nggj-tindasa-rikin-av-vitlki), 2026-04-23
- [Deployment] Vatican City · Interpreter — 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. [catholicworldreport.com](https://www.catholicworldreport.com/2026/02/20/vatican-to-use-ai-to-translate-masses-at-st-peters-into-60-languages-%E2%80%8B%E2%80%8Bin-real-time/), 2026-02-20
- [Deployment] Cuba · Television presenter — 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. [perlavision.cu](https://www.perlavision.cu/cuerdaviva-tvcubana-ia/), 2026-04-28
- [Deployment] Curaçao · Filmmaker — 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. [curacaochronicle.com](https://www.curacaochronicle.com/post/unknown/curacao-innovation-hub-takes-second-place-worldwide-at-ai-film-festival), 2026-08-27
- [Deployment] Hungary · Video editor — 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. [index.hu](https://index.hu/techtud/2026/09/07/aisummit2026-ai-summit-mesterseges-intelligencia-rtl-verebely-marcell-televizios-musorgyartas), 2026-09-07
- [Deployment] Kyrgyzstan · news anchor — 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. [open.kg](https://open.kg/en/news/exclusive/77477-nejroseti-v-parlamente-rabotu-zhogorku-kenesha-budut-osveschat-ii-veduschie.html), 2026-03-02
- [Deployment] Czechia · Customer service adviser — 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. [protext.cz](https://www.protext.cz/zpravy/show/58999), 4 Sep 2026
- [Deployment] Somalia · media producer — 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. [iafrica.com](https://iafrica.com/somalia-presents-ai-adoption-plan-at-east-africa-summit-positions-itself-as-regional-digital-leader/), 2026-04-13
- [Regulation] South Sudan — 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. [sudanspost.com](https://www.sudanspost.com/media-authority-gives-green-light-for-ai-adoption-in-newsrooms/), 2026-06-02

## Transport & Logistics

- [Deployment] Singapore · Airline operations — 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. [techwireasia.com](https://techwireasia.com/2026/08/singapore-airlines-ai-applications/), 18 Aug 2026
- [Deployment] United States · Revenue / pricing — 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. [thetravel.com](https://www.thetravel.com/delta-air-lines-to-ramp-up-ai-use-to-raise-ticket-prices-and-cut-costs/), 19–21 Aug 2026
- [Deployment] Singapore · Ship crew — 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. [thedigitalship.com](https://thedigitalship.com/news/electronics-navigation/eastern-pacific-shipping-expands-orca-ai-rollout-after-48-drop-in-close-encounters/), 26 Aug 2026
- [Deployment] Japan · Marine surveyor — 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. [marinelink.com](https://www.marinelink.com/news/classnk-launches-survey-compass-ai-542431), 25 Aug 2026
- [Deployment] Netherlands · Back-office ops — 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. [pulse2.com](https://pulse2.com/postnl-ai-agents-and-parcel-lockers-support-expanded-e180-million-cost-savings-plan/), 2026-08-03
- [Deployment] Ireland · All staff (airline) — 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. [unite.ai](https://www.unite.ai/ryanair-adopts-gemini-enterprise-across-crew-logistics-and-flight-operations/), 2026-08-12
- [Deployment] Panama · airport passenger services agent — 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. [infobae.com](https://www.infobae.com/panama/2026/08/22/aeropuerto-de-tocumen-en-panama-incorporara-biometria-y-marketplace-a-su-plataforma-digital/), 2026-08-21
- [Deployment] United States · transit customer service agent — 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. [roi-nj.com](https://www.roi-nj.com/2026/09/01/transportation/nj-transit-debuts-digital-upgrades-including-ai-powered-chatbot-navvie/), 2026-09-01
- [Deployment] Canada · operations controller — 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. [avitrader.com](https://avitrader.com/2026/08/20/air-transat-deploys-stratosx-x360-platform-becomes-launch-costumer/), 2026-08-20
- [Deployment] South Korea · Warehouse packer — 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. [koreatimes.co.kr](https://www.koreatimes.co.kr/business/companies/20260903/cj-logistics-deploys-humanoid-robots-at-warehouses-in-korea), 2026-09-03
- [Deployment] Germany · Warehouse picker — 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. [corporate.zalando.com](https://corporate.zalando.com/en/technology/zalando-install-50-ai-powered-nomagic-robots-its-european-fulfilment-centres), 2026-03-19
- [Deployment] United Kingdom · Private hire driver — 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. [wayve.ai](https://wayve.ai/press/wayve-uber-launch-autonomous-rides/), 2026-09-03
- [Deployment] United Arab Emirates · Bus driver — 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. [sharjah24.ae](https://sharjah24.ae/en/Articles/2026/08/20/mm05), 2026-08-20
- [Deployment] Ethiopia · Freight forwarding agent — 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. [ethiopiatoday.net](https://ethiopiatoday.net/ethiopian-shipping-logistics-launches-ai-platform-to-automate-freight-operations/), 2026-09-04
- [Deployment] Dominican Republic · Airport security screener — 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. [diariolibre.com](https://www.diariolibre.com/economia/turismo/2026/08/19/inteligencia-artificial-en-aeropuerto-punta-cana-para-deteccion-de-con/3633858), 2026-08-19
- [Deployment] Lithuania · Customer service agent — 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. [madeinvilnius.lt](https://madeinvilnius.lt/transportas/viesasis-transportas/vilniaus-viesasis-transportas-zengia-i-nauja-etapa-dali-darbo-perims-dirbtinis-intelektas/), 2026-06-14
- [Deployment] Kenya · Airline pricing analyst — 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. [hapakenya.com](https://hapakenya.com/2026/05/19/how-kenya-airways-is-using-ai-to-offer-competitive-ticket-pricing-to-customers/), 2026-05-19
- [Deployment] South Africa · Road maintenance engineer — 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. [bentley.com](https://www.bentley.com/news/western-cape-department-of-infrastructure-partners-with-bentley-systems-to-modernize-roadway-safety-in-south-africa/), 2026-05-19
- [Deployment] Mozambique · Container terminal operator — 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. [mznews.co.mz](https://mznews.co.mz/cornelder-introduz-sistema-de-inteligencia-artificial-para-manuseamento-de-carga-no-porto-da-beira/), 2026-01-02
- [Deployment] Haiti · Airline customer-service agent — 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. [juno7.ht](https://www.juno7.ht/sunrise-airways-assistant-whatsapp-intelligence-artificielle/), 2026-07-24
- [Deployment] United Arab Emirates · Bus operations controller — 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. [khaleejtimes.com](https://www.khaleejtimes.com/uae/transport/rta-dubai-bus-deployment-ai-system-launched), 2026-07-06
- [Deployment] Oman · Road paving crew — 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. [omanobserver.om](https://www.omanobserver.om/article/1189957/oman/community/transport-ministry-launches-omans-first-ai-asphalt-paving-tech-in-dhofar), 2026-05-20
- [Deployment] United States · Airport customer service agent — 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. [newjersey.news12.com](https://newjersey.news12.com/2026/07/16/port-authority-launches-ai-travel-assistant-at-jfk-newark-laguardia-and-stewart-airports/6WvmxbIgGysEFkasTHhQan), 2026-07-16
- [Deployment] Croatia · Taxi driver — 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.' [zagdaily.com](https://zagdaily.com/connected/croatia-launches-europes-first-commercial-robotaxi-service/), 2026-04-08
- [Deployment] Croatia · Autonomous vehicle safety operator — 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. [prnewswire.com](http://www.prnewswire.com/news-releases/ponyai-and-verne-kick-off-fully-driverless-robotaxi-test-rides-in-zagreb-302874840.html), 2026-09-10
- [Deployment] Australia · Traffic signal engineer — 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. [carsguide.com.au](https://www.carsguide.com.au/car-news/new-ai-traffic-lights-trial-in-moreton-bay-queensland-to-drive-biggest-change-to-traffic), 2026-08-27
- [Deployment] Australia · Traffic signal engineer — 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. [abc.net.au](https://www.abc.net.au/news/2026-09-10/brisbane-traffic-light-signalling-artificial-intelligence-pilot/107136566), 2026-09-10
- [Deployment] Malaysia · Highway traffic control operator — 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. [paultan.org](https://paultan.org/2026/09/08/ijm-smart-highway-traffic-control-centre-launched/), 2026-09-08
- [Deployment] Singapore · Warehouse operative — 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. [container-news.com](https://container-news.com/ceva-logistics-pilots-autonomous-electric-vehicles-at-singapore-hub/), 2026-09-14
- [Deployment] South Korea · Logistics safety manager — 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. [en.sedaily.com](https://en.sedaily.com/finance/2026/09/11/hanjin-rolls-out-ai-cameras-to-flag-safety-risks-at), 2026-09-11
- [Deployment] Malaysia · Logistics operator — 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. [truckandbusnews.net](https://www.truckandbusnews.net/pos-malaysia-first-autonomous-logistics-vehicle/), 2026-01-22
- [Deployment] Switzerland · delivery driver — 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. [swisstrans.info](https://swisstrans.info/en/2026/05/21/planzer-loxo-launch-automated-city-logistics-in-bern/), 2026-05-21
- [Deployment] Tajikistan · Transport operators — 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. [en.avesta.tj](https://en.avesta.tj/2026/06/17/tajikistans-ministry-of-transport-has-launched-an-ai-call-center-for-carriers/), 2026-06-17
- [Deployment] Greenland · Ship captain / navigator / seafarer — 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. [sermitsiaq.ag](https://www.sermitsiaq.ag/samfund/gronlandske-sofolk-har-testet-iskort-baseret-pa-ai/2355475), 2026-03-11
- [Deployment] Malta · Bus driver — 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. [publictransport.com.mt](https://www.publictransport.com.mt/autonomous-bus/), 2026-08-19

## Manufacturing, Energy & Resources

- [Deployment] United Arab Emirates · Rig / field operator — 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. [worldoil.com](https://worldoil.com/news/2026/8/4/adnoc-slb-deploy-ai-platform-across-more-than-120-drilling-rigs/), 4 Aug 2026
- [Deployment] United States · Construction / field — 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. [constructiondive.com](https://constructiondive.com/news/bedrock-robotics-fully-autonomous-excavators-jobsites/828267/), 19 Aug 2026
- [Deployment] United States · Construction / field — 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. [constructiondive.com](https://constructiondive.com/news/novo-construction-buildcheck-drawing-diffs-ai/826758/), 5 Aug 2026
- [Deployment] United States · Operations / field — 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. [techcrunch.com](https://techcrunch.com/2026/08/30/caterpillar-is-bringing-to-ai-deployment-what-it-learned-from-automating-mining/), 30 Aug 2026
- [Deployment] China · Haulage operator — 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. [im-mining.com](https://im-mining.com/2026/08/13/cidis-metamine-ahs-deployment-grows-to-1900-mining-trucks-at-nearly-40-mines/), 13 Aug 2026
- [Deployment] South Africa · Mining operator — 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. [im-mining.com](https://im-mining.com/2026/08/30/bme-reimagines-mining-by-turning-data-into-value/), 30 Aug 2026
- [Deployment] United States · Construction / field — 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. [constructiondive.com](https://constructiondive.com/news/construction-pros-jobsite-technology-ai-safety/827678/), 12 Aug 2026
- [Evidence of adoption] Global — 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. [rics.org](https://www.rics.org/news-insights/ai-in-commercial-property-and-construction-report-2026), 2026
- [Evidence of adoption] United States — 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. [constructiondive.com](https://www.constructiondive.com/news/builders-ai-transform-businesses-survey/807555/), 2025
- [Evidence of impact] Global — AI could automate 39% of nonphysical work in construction and 50% in architecture and engineering, McKinsey estimates. This is automation potential, not current adoption. [mckinsey.com](https://www.mckinsey.com/industries/engineering-construction-and-building-materials/our-insights/how-ai-is-reshaping-the-future-of-the-aec-industry), 
- [Deployment] United States · Construction / field — 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. [constructiondive.com](https://www.constructiondive.com/news/zachry-construction-ai-estimation-alice/742951/), 
- [Deployment] Brazil · processing plant operator — 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. [im-mining.com](https://im-mining.com/2026/08/12/vale-and-abb-sign-partnership-to-accelerate-digital-transformation-in-iron-ore-operations-in-brazil/), 2026-08-12
- [Deployment] Guyana · exploration geoscientist — 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. [worldoil.com](https://www.worldoil.com/news/2026/8/13/exxonmobil-identifies-four-guyana-exploration-opportunities-using-ai/), 2026-08-13
- [Deployment] China · Drilling site safety supervisor — 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. [ecns.cn](http://www.ecns.cn/cns-wire/2026-09-02/detail-ihfiwhfw6700410.shtml), 2026-09-02
- [Deployment] Jordan · Water network operator — 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. [petra.gov.jo](https://petra.gov.jo/dotcleen/index.php/en/news/water-digital-economy-ministers-review-ai-water-project), 2026-08-09
- [Deployment] Canada · Assembly-plant material handler — 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. [therobotreport.com](https://www.therobotreport.com/toyota-motor-manufacturing-canada-deploys-agility-robotics-digit-humanoids/), 2026-02-19
- [Deployment] Botswana · Exploration geologist — 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. [investegate.co.uk](https://www.investegate.co.uk/announcement/rns/botswana-minerals-plc--bmin/ai-exploration-targets-defined-sampling-to-begin/9755171), 2026-09-04
- [Evidence of impact] Chile — 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. [elboyaldia.cl](https://www.elboyaldia.cl/noticia/economia/codelco-alcanzo-un-40-de-mejora-en-la-produccion-gracias-la-incorporacion-de-la-int), 2026-09-09
- [Evidence of adoption] Costa Rica — 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. [cicr.com](https://cicr.com/cicr-uso-de-la-inteligencia-artificial-gana-terreno-en-la-industria-manufacturera-pero-persisten-brechas-en-su-adopcion/), 2026-09-07
- [Deployment] Colombia · Production and supply-chain planner — 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. [elnorte.com.co](https://elnorte.com.co/cementos-argos-revoluciona-su-operacion-en-cartagena-con-el-poder-de-la-inteligencia-artificial/), 2026-06-17
- [Deployment] Peru · Mine operations controller — 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. [desdeadentro.pe](https://www.desdeadentro.pe/2026/02/minera-chinalco-culmina-la-implementacion-de-su-centro-de-gestion-integrada-de-operaciones-gio/), 2026-02-19
- [Evidence of impact] Chile — 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. [redimin.cl](https://www.redimin.cl/codelco-el-teniente-reduce-90-conductas-de-riesgo-con-inteligencia-artificial-y-40-camaras-en-faenas), 2026-04-02
- [Deployment] Saudi Arabia · Water resource manager — 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. [tradearabia.com](https://tradearabia.com/News/388776/Saudi-Water-Authority-IBM-launch-AI-powered-Digital-H2O-platform), 2026-02-03
- [Deployment] Saudi Arabia · Refinery planning engineer — 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. [emerson.com](https://www.emerson.com/en/corporate/news/2026/emerson-aramco-deploy-ai-solution-higher-refinery-yield-volume), 2026-04-28
- [Evidence of adoption] United Arab Emirates — 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. [khaleejtimes.com](https://www.khaleejtimes.com/uae/uae-agentic-ai-agents-public-services-retreat), 2026-05-20
- [Deployment] Libya · Directional driller — 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. [attaqa.net](https://attaqa.net/2026/03/27/%D8%AD%D9%81%D8%B1-%D8%A3%D9%88%D9%84-%D8%A8%D8%A6%D8%B1-%D8%A8%D8%A7%D8%B3%D8%AA%D8%B9%D9%85%D8%A7%D9%84-%D8%A7%D9%84%D8%B0%D9%83%D8%A7%D8%A1-%D8%A7%D9%84%D8%A7%D8%B5%D8%B7%D9%86%D8%A7%D8%B9%D9%8A/), 2026-03-27
- [Deployment] Turkey · Call-centre agent — 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. [suvecevre.com](https://www.suvecevre.com/yayin/327/-iski-ai-yapay-zeka-projesi-tanitildi_40363.html), 2026-07-22
- [Deployment] United States · utility load forecaster — 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. [utilitydive.com](https://www.utilitydive.com/news/santee-cooper-partners-with-google-in-shift-toward-ai-driven-forecasting/829564/), 2026-09-03
- [Deployment] United States · nuclear engineer — 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. [nuscalepower.com](https://www.nuscalepower.com/press-releases/2026/nuscale-power-deploys-nuclear-specific-ai-from-npx-and-nuclearn-to-speed-development-of-its-advanced-reactor-program), 2026-08-25
- [Deployment] Belarus · Designer — 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. [belta.by](https://belta.by/society/view/produktsija-s-ispolzovaniem-ii-uzhe-prodaetsja-kak-tsifrovizatsija-menjaet-dizajn-belorusskoj-odezhdy-766979-2026/), 2026-02-28
- [Deployment] Azerbaijan · Plant operator — 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%. [weforum.org](https://www.weforum.org/press/2026/01/global-lighthouse-network-recognizes-23-new-sites-launches-ai-platform-for-industrial-transformation/), 2026-01-15
- [Deployment] Burundi · Exploration geologist — 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. [tech.africa](https://tech.africa/kobold-metals-ai-mining-africa/), 2026-06-20
- [Deployment] Czechia · Hydropower maintenance planner — 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. [cez.cz](https://www.cez.cz/cs/pro-media/tiskove-zpravy/stechovice-nasly-poklad-v-umele-inteligenci.-ai-vcas-odhali-hrozici-poruchy-v-precerpavaci-elektrarne-232583), 2026-03-18
- [Deployment] Cuba · Electricity dispatcher — 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. [radiorebelde.cu](https://www.radiorebelde.cu/chequea-presidente-diaz-canel-situacion-electrica-de-la-habana-07082026/), 2026-08-07
- [Deployment] DR Congo · Exploration geologist — 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. [mining.com](https://www.mining.com/kobold-launches-ai-driven-lithium-push-in-drc/), 2026-04-13
- [Deployment] China · Exploration geologist — 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. [english.news.cn](https://english.news.cn/20260911/c523e91f55e842c4acdcf1e5826ff0fe/c.html), 2026-09-11
- [Deployment] South Korea · Production engineer — 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. [en.sedaily.com](https://en.sedaily.com/finance/2026/09/13/lg-display-expands-ai-program-letting-non-coders-build-apps), 2026-09-13
- [Deployment] Latvia · grid inspection engineer — 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. [kursors.lv](https://kursors.lv/2026/06/19/as-augstsprieguma-tikls-ar-maksligo-intelektu-uzlabos-parvades-tikla-infrastrukturas-uzraudzibu/), 2026-06-19
- [Deployment] Turkmenistan · Petroleum engineer — 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. [newscentralasia.net](https://www.newscentralasia.net/2026/06/27/turkmenistan-advances-digitalization-and-ai-in-oil-and-gas-sector/), 2026-06-27
- [Deployment] Algeria · Drilling rig operator — 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. [jeune-independant.net](https://www.jeune-independant.net/daoudi-a-lentp-de-hassi-messaoud-lia-pour-optimiser-les-operations-de-forage/), 2025-12-22

## Agriculture & Food

- [Deployment] Nigeria · Agriculture officer — 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. [spaceinafrica.com](https://spaceinafrica.com/2026/07/21/nigeria-to-deploy-ai-and-satellite-powered-crop-monitoring-system/), 2026-07-25
- [Evidence of adoption] United States — 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. [gao.gov](https://www.gao.gov/products/gao-24-105962), 2023
- [Evidence of adoption] United States — 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. [ers.usda.gov](https://www.ers.usda.gov/data-products/charts-of-note/110550), 2023
- [Evidence of impact] Global — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC11820920/), 2025
- [Deployment] Kenya · Farmer — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC7775399/), 
- [Evidence of impact] United States — 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. [ers.usda.gov](https://www.ers.usda.gov/publications/113704), 2021
- [Evidence of adoption] United States — 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. [ers.usda.gov](https://www.ers.usda.gov/publications/113704), 2021
- [Evidence of impact] United States — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC12268811/), 2024
- [Regulation] European Union — 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. [digital-strategy.ec.europa.eu](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), 2026
- [Evidence of adoption] United States — 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. [ers.usda.gov](https://www.ers.usda.gov/data-products/charts-of-note/110550), 2023
- [Deployment] Pakistan · Farmer — 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. [app.com.pk](https://www.app.com.pk/national/telenor-launches-pakistans-first-ai-powered-agriculture-voice-bot-kissan-dost-bashir/), 2026-04-23
- [Deployment] India · Farmer — 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. [tribuneindia.com](https://www.tribuneindia.com/news/india/shivraj-chouhan-launches-ai-powered-bharat-vistaar-platform-for-farmers/), 2026-02-17
- [Deployment] Brazil · Farmer — 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. [cnnbrasil.com.br](https://www.cnnbrasil.com.br/agro/cna-lanca-ferramenta-de-inteligencia-artificial-para-apoiar-produtores/), 2026-08-25
- [Regulation] Namibia — 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. [namibian.com.na](https://www.namibian.com.na/cabinet-blocks-us-ai-crop-monitoring-tender/), 2026-08-20
- [Deployment] Rwanda · Agricultural extension officer — 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. [ktpress.rw](https://www.ktpress.rw/2026/09/here-is-our-ai-plan-to-reach-2-5-million-rwandan-farmers/), 2026-09-03
- [Deployment] Togo · Agricultural extension officer — 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. [lacinquieme.tg](https://www.lacinquieme.tg/agriculture-dans-les-savanes-le-togo-veut-voir-venir-les-risques-climatiques-avant-quils-ne-frappent-les-cultures/), 2026-09-07
- [Deployment] Saudi Arabia · Procurement and supply-chain planner — 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. [millingmea.com](https://millingmea.com/first-mills-deploys-ai-across-production-facilities-to-transform-milling-operations/), 2026-04-14
- [Deployment] Albania · Farmer — 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. [rtsh.al](https://rtsh.al/rti/en/virtual-farming-assistant-launched-at-digital-agriculture-conference/), 2026-06-01
- [Deployment] Andorra · Forest technician — 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. [alto.ad](https://www.alto.ad/environment/2026/07/andorra-3d-scanning-drones-mountain-forests), 2026-07-18
- [Evidence of impact] Benin — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC13246710/), 2026-05-26
- [Evidence of impact] Syria — 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. [pmc.ncbi.nlm.nih.gov](https://pmc.ncbi.nlm.nih.gov/articles/PMC13196713/), 2026-05-19
- [Deployment] Azerbaijan · Farmer — 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. [trend.az](https://www.trend.az/business/4221738.html), 2026-09-08
- [Deployment] Azerbaijan · Food safety inspector — 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. [az.trend.az](https://az.trend.az/azerbaijan/society/4222748.html), 2026-09-10
- [Evidence of impact] British Indian Ocean Territory — 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. [doi.org](https://doi.org/10.1098/rsif.2025.0172), 2026-05-01
- [Deployment] Cambodia · Conservation biologist — 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. [phys.org](https://phys.org/news/2026-06-secret-cameras-mics-ai-reveal.html), 2026-06-25
- [Deployment] Faroe Islands · Feed centre operator — 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. [kvf.fo](https://kvf.fo/netvarp/sv/2026/05/22/nyggjur-fodurflaki-vid-vitliki), 2026-05-22
- [Deployment] Cayman Islands · Marine biologist — 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. [caymancompass.com](https://www.caymancompass.com/2026/01/20/modern-technology-helps-protect-nassau-grouper-maternity-wards/), 2026-01-20
- [Deployment] Fiji · farmer — 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. [fijisun.com.fj](https://fijisun.com.fj/news/nation/digital-push-to-future-proof-fijis-sugar-industry), 2026-09-04
- [Deployment] Kyrgyzstan · livestock farmer — 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. [distantnews.com](https://distantnews.com/article/koica-presents-ai-based-livestock-health-monitoring-system-t-bqv517kc), 2026-08-11
- [Deployment] Slovakia · hop grower — 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. [hnonline.sk](https://hnonline.sk/hn-nativ/96276191-poslednej-slovenskej-chmelnici-radi-umela-inteligencia-v-nemsovej-osadili-specialne-meteostanice), 2026-04-17
- [Deployment] Vanuatu · Forestry officer — 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. [spc.int](https://www.spc.int/updates/blog/blog-post/2026/08/vanuatu-puts-invasive-plant-mapping-to-the-test), 2026-08-21
- [Deployment] Zimbabwe · farmer — 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. [technomag.co.zw](https://technomag.co.zw/ecofarmer-ai-to-boost-productivity-for-zimbabwean-farmers/), 2026-04-21
- [Deployment] Somalia · farmer — 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. [bbc.com](https://www.bbc.com/somali/articles/cn0ewgk40ggo), 2026-02-19
- [Deployment] Canada · Head grower — 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. [blue-radix.com](https://blue-radix.com/projects/jpi-farms/), 2025-09-16

## Technology & Software

- [Evidence of adoption] Global — 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. [survey.stackoverflow.co](https://survey.stackoverflow.co/2025/ai/), 2025
- [Evidence of adoption] Global — 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. [cloud.google.com](https://cloud.google.com/blog/products/ai-machine-learning/announcing-the-2025-dora-report), 2025
- [Evidence of impact] Global — 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. [survey.stackoverflow.co](https://survey.stackoverflow.co/2025/ai/), 2025
- [Deployment] United States · Software developer — 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'. [blog.google](https://blog.google/innovation-and-ai/infrastructure-and-cloud/google-cloud/cloud-next-2026-sundar-pichai/), 2024
- [Deployment] United States · Regulator — 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. [ftc.gov](https://www.ftc.gov/news-events/news/press-releases/2025/09/ftc-launches-inquiry-ai-chatbots-acting-companions), 2026
- [Evidence of impact] Global — 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. [veracode.com](https://www.veracode.com/blog/2026-genai-code-security-report-ai-risk/), 2026
- [Evidence of impact] Global — 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. [dora.dev](https://dora.dev/ai/roi/report/), 2026
- [Evidence of impact] Global — 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. [metr.org](https://metr.org/blog/2026-02-24-uplift-update/), 2026
- [Regulation] United States — 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. [morganlewis.com](https://www.morganlewis.com/pubs/2026/08/new-california-ai-disclosure-rules-become-operative), 2026-08-04
- [Regulation] United States — 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. [consumerfinancemonitor.com](https://www.consumerfinancemonitor.com/2026/08/12/colorado-publishes-draft-admt-regulations/), 2026-08-12
- [Regulation] Japan — 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. [japantoday.com](https://japantoday.com/category/tech/japan-adopts-ip-use-rules-for-ai-firms-urges-training-data-disclosure), 2026-08-26
- [Regulation] Vietnam — 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. [bakermckenzie.com](https://www.bakermckenzie.com/en/insight/publications/2026/02/vietnam-artificial-intelligence-law-foundation-and-outlook), 2026-02-24
- [Regulation] Singapore — 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. [mddi.gov.sg](https://www.mddi.gov.sg/newsroom/singapore-launches-new-model-ai-governance-framework-for-agentic-ai--/), 2026-01-22
- [Deployment] Brazil · Software engineer — 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. [tiinside.com.br](https://tiinside.com.br/01/09/2026/b3-adota-inteligencia-artificial-para-desenvolver-softwares-e-reduz-em-35-tempo-de-entrega/), 2026-09-01
- [Evidence of adoption] Uruguay — 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. [subrayado.com.uy](https://www.subrayado.com.uy/los-uruguayos-miran-mas-preocupacion-que-entusiasmo-el-creciente-uso-la-inteligencia-artificial-n1015896), 2026-08-20
- [Regulation] United States — 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. [gov.ca.gov](http://www.gov.ca.gov/2026/09/09/governor-newsom-signs-first-in-the-nation-ai-safeguards-to-protect-californians-calls-on-the-federal-government-to-do-its-part/), 2026-09-09
- [Evidence of adoption] Afghanistan — 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. [pajhwok.com](https://pajhwok.com/2026/05/04/survey-finds-widespread-ai-use-among-afghan-youth/), 2026-05-04
- [Evidence of adoption] Moldova — 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. [moldova1.md](https://moldova1.md/p/82184), 2026-08-02
- [Evidence of adoption] Hong Kong — 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. [hongkongfp.com](https://hongkongfp.com/2026/06/03/over-90-of-hong-kong-organisations-use-ai-tools-despite-training-and-policy-gaps-survey-finds/), 2026-06-03
- [Evidence of adoption] North Macedonia — 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. [new.mia.mk](https://new.mia.mk/al/story/ai-used-by-45.4-of-young-people-in-the-country-63.8-in-the-eu), 2026-02-11
- [Evidence of adoption] Russia — 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. [noah-news.com](https://noah-news.com/russian-companies-rapidly-adopt-ai-but-cautious-about-autonomous-systems-study-s/), 2026-08-25
- [Evidence of adoption] Slovenia — 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. [en.doba.si](https://www.en.doba.si/doba-ai-maturity-index), 2026-08-04
- [Evidence of adoption] Réunion — 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. [lequotidien.re](https://www.lequotidien.re/article/societe/2026/08/24/la-reunion-nechappe-pas-a-la-vague-ia), 2026-08-24

## Business Operations

- [Deployment] United States · All staff — 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. [pymnts.com](https://pymnts.com/news/artificial-intelligence/2026/cisco-deploys-custom-ai-agent-to-entire-90000-person-workforce), 27 Aug 2026
- [Deployment] Netherlands · Compliance / onboarding — 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. [fintech.global](https://fintech.global/2026/08/14/ai-agents-move-from-pilot-to-workforce-in-bank-compliance/), 14 Aug 2026
- [Deployment] Japan · Operations staff — 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. [asia.nikkei.com](https://asia.nikkei.com/business/technology/artificial-intelligence/ai-agents-are-the-only-employees-at-japan-tech-group-nec-s-new-department), 2026-08-11
- [Deployment] Germany · All staff — 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. [technologyrecord.com](https://www.technologyrecord.com/article/sportfive-scales-microsoft-365-copilot-to-93-per-cent-adoption), 2026-08-27
- [Evidence of adoption] Global — 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. [mckinsey.com](https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai), 2026
- [Evidence of adoption] United States — 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. [census.gov](https://www.census.gov/hfp/btos/data_downloads), 2026
- [Evidence of adoption] Global — 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. [hai.stanford.edu](https://hai.stanford.edu/news/inside-the-ai-index-12-takeaways-from-the-2026-report), 2025
- [Deployment] Singapore · Operations staff — 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. [dbs.com](https://www.dbs.com/annualreports/2025/letter-from-chairman-ceo.html), 2025
- [Evidence of adoption] United States — 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. [census.gov](https://www.census.gov/library/working-papers/2026/adrm/CES-WP-26-25.html), 2026
- [Evidence of adoption] United Kingdom — 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. [ons.gov.uk](https://www.ons.gov.uk/businessindustryandtrade/business/businessservices/articles/artificialintelligenceinukbusinesses/2023to2026), 2026
- [Evidence of adoption] European Union — 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. [ec.europa.eu](https://ec.europa.eu/eurostat/web/products-eurostat-news/w/ddn-20251211-2), 2025
- [Evidence of adoption] Canada — 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. [www150.statcan.gc.ca](https://www150.statcan.gc.ca/n1/pub/11-621-m/11-621-m2026010-eng.htm), 2026
- [Evidence of adoption] Global — 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%. [nber.org](https://www.nber.org/papers/w34836), 2026
- [Evidence of impact] Global — 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. [nber.org](https://www.nber.org/papers/w33795), 2025
- [Evidence of adoption] Canada — 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. [www150.statcan.gc.ca](https://www150.statcan.gc.ca/n1/pub/36-28-0001/2026004/article/00002-eng.htm), 2021
- [Evidence of adoption] United States — 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. [census.gov](https://www.census.gov/library/stories/2026/08/ai-use-at-work.html), 2026
- [Regulation] European Union — 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. [digital-strategy.ec.europa.eu](https://digital-strategy.ec.europa.eu/en/news/ai-omnibus-enters-force), 2026
- [Evidence of adoption] Chile — 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. [gerencia.cl](https://www.gerencia.cl/inteligencia-artificial/estudio-revela-que-la-ia-ya-es-parte-del-88-de-las-empresas-chilenas-pero-se-implementa-sin-una-estrategia-clara/), 2026-08-24
- [Evidence of adoption] Colombia — 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. [colombiaone.com](https://colombiaone.com/2026/08/31/ai-adoption-colombia/), 2026-08-31
- [Evidence of adoption] Thailand — 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. [retailnews.asia](https://retailnews.asia/world-bank-urges-thailand-to-lift-12-ai-adoption-rate-for-2037-goal), 2026-09-06
- [Deployment] United Arab Emirates · Civil servant — 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. [mediaoffice.abudhabi](https://www.mediaoffice.abudhabi/en/technology/abu-dhabi-government-partners-with-microsoft-for-frontier-employee-programme-in-one-of-public-sectors-largest-ai-productivity-rollouts/), 2026-07-06
- [Deployment] South Africa · Bank operations analyst — 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. [it-online.co.za](https://it-online.co.za/2026/06/25/investec-in-sa-first-with-copilot-rollout/), 2026-06
- [Evidence of adoption] Egypt — 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. [egyin.com](https://www.egyin.com/article/145549/), 2026-09-07
- [Deployment] Japan · Office worker — 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. [prtimes.jp](https://prtimes.jp/main/html/rd/p/000000085.000159136.html), 2025-12-26
- [Evidence of adoption] Taiwan — 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. [aif.tw](https://aif.tw/event/ai-research/), 2026-03-18
- [Evidence of impact] Italy — 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. [bancaditalia.it](https://www.bancaditalia.it/pubblicazioni/qef/2026-1005/index.html), 2026-03
- [Evidence of adoption] Puerto Rico — 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. [newsismybusiness.com](https://newsismybusiness.com/es/los-emprendedores-puertorriquenos-adoptan-la-ia-y-luchan-con-el-capital/), 2026-09-07
- [Deployment] Brazil · Public audit court staff — 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. [tce.pr.gov.br](https://www.tce.pr.gov.br/noticias/tce-pr-adota-ferramenta-de-ia-e-abre-trilha-de-capacitacao-sobre-o-assunto-a-servidores.htm), 2026-02-24
- [Deployment] Qatar · Bank staff — 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. [thepeninsulaqatar.com](https://thepeninsulaqatar.com/article/21/05/2026/alrayan-bank-adopts-microsoft-365-copilot-to-enhance-innovation-and-enterprise-productivity), 2026-05-21
- [Evidence of impact] United States — 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. [bamboohr.com](https://www.bamboohr.com/about-bamboohr/press-release/bamboohr-research-redesigning-work-ai-performance-review), 2026-09-01
- [Evidence of adoption] Canada — 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. [news.sap.com](https://news.sap.com/canada/2026/09/canadian-businesses-race-to-adopt-agentic-ai-but-97-say-they-arent-fully-ready/), 2026-09-10
- [Evidence of adoption] Lebanon — 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. [byblos.byblosbank.com](https://byblos.byblosbank.com/files/Library/Assets/Gallery/Publications/LebanonThisWeek/Lebanon%20this%20Week%20912.pdf), 2026-06-06
- [Evidence of adoption] Albania — 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. [bulevardnews.com](https://www.bulevardnews.com/2026/09/08/ibrahimaj-vetem-9-e-ndermarrjeve-shqiptare-perdorin-ia-duhet-te-kapim-ritmin/), 2026-09-08
- [Evidence of adoption] Andorra — 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. [alto.ad](https://www.alto.ad/business/2026/08/andorra-digitalisation-programme-closed-budget-a), 2026-08-04
- [Evidence of adoption] Andorra — 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. [alto.ad](https://www.alto.ad/business/2026/08/andorra-digital-divide-elderly-smes-skills), 2026-08-13
- [Evidence of adoption] Belarus — 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. [belretail.by](https://belretail.by/article/chto-dumayut-belarusyi-o-razvitii-ii), 2026-01-29
- [Evidence of adoption] Bosnia and Herzegovina — 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. [microsoft.com](https://www.microsoft.com/en-us/research/wp-content/uploads/2026/05/Microsoft-AI-Diffusion-Report-2026-Q1.pdf), 2026-05-07
- [Evidence of adoption] Czechia — 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. [itbiz.cz](https://www.itbiz.cz/ceske-firmy-vetsinove-tvrdi-ze-jim-ai-pomaha-jeji-vyuzivani-se-lisi-podle-oboru/), 2026-06-24
- [Evidence of adoption] Bermuda — 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. [bernews.com](https://bernews.com/2026/01/survey-artificial-intelligence-use-in-bermuda/), 2026-01-31
- [Evidence of adoption] Estonia — 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. [kpmg.com](https://kpmg.com/ee/et/insights/2026/04/turvalisuse-nimel--pooled-eesti-suurfirmad-usaldavad-vaid-lokaal.html), 2026-04-07
- [Evidence of adoption] Central African Republic — 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.com](https://www.microsoft.com/en-us/research/wp-content/uploads/2026/05/Microsoft-AI-Diffusion-Report-2026-Q1.pdf), 2026-05-07
- [Evidence of adoption] Chad — 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. [microsoft.com](https://www.microsoft.com/en-us/research/wp-content/uploads/2026/05/Microsoft-AI-Diffusion-Report-2026-Q1.pdf), 2026-05
- [Evidence of adoption] DR Congo — 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. [frontiersin.org](https://www.frontiersin.org/journals/artificial-intelligence/articles/10.3389/frai.2026.1868452/full), 2026-06-03
- [Evidence of adoption] Cyprus — 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. [tothemaonline.com](https://www.tothemaonline.com/Article/471779/poses-kypriakes-epiheirhseis-hrhsimopoioyn-ai-ti-deihnoyn-oi-arithmoi---to-hasma-kai-to-neo-stoihhma---), 2026-09-06
- [Deployment] Hungary · Small business owner — 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. [economx.hu](https://www.economx.hu/mesterseges-intelligencia/2026/03/18/mesterseges-intelligencia-kkv-palkovics-laszlo-ai-start-500-program-825635/), 2026-03-18
- [Evidence of adoption] Hungary — 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. [hirek.prim.hu](https://hirek.prim.hu/cikk/2026/07/13/egyre_tobb_ceg_epiti_be_mukodesebe_az_ai-t_rekordot_mutat_a_k_h_felmerese), 2026-07-13
- [Evidence of adoption] Thailand — 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. [nectec.or.th](https://www.nectec.or.th/news/news-pr-news/ai-readiness-2026.html), 2026-09-03
- [Evidence of adoption] Thailand — 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. [mgronline.com](https://mgronline.com/politics/detail/9690000088562), 2026-09-10
- [Evidence of adoption] Iceland — 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. [statice.is](https://statice.is/publications/news-archive/science-and-technology/information-and-communication-technology-2024/), 2026-03-10
- [Evidence of adoption] Georgia — 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. [ai-news.ge](https://ai-news.ge/en/article/bag-survey-84-percent-georgian-companies-adopting-ai), 2026-07-20
- [Evidence of adoption] Martinique — 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. [awitec.fr](https://www.awitec.fr/blog/barometre-ia-martinique-2026-usages-chiffres-intelligence-artificielle-entreprises/), 2026-06-11
- [Evidence of adoption] France — 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. [presse.bpifrance.fr](https://presse.bpifrance.fr/un-an-du-plan-osez-lia-un-premier-bilan-positif-et-une-nouvelle-etape-pour-accelerer-ladoption-de-lia-dans-les-entreprises/?lang=fra), 8 Sep 2026
- [Evidence of adoption] Austria — 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. [presse-nachrichten.de](https://www.presse-nachrichten.de/2026/09/09/schatten-ki-in-oesterreich-jeder-sechste-nutzt-im-job-heimlich-ki-assistenten-42-verzichten-ganz-auf-ki/), 9 Sep 2026
- [Evidence of adoption] Greenland — 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. [ilik.gl](https://ilik.gl/kunstig-intelligens-vinder-frem-i-groenlandske-virksomheder/), 2026-01-12
- [Deployment] Mayotte · business advisor — 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. [inforadar.fr](https://inforadar.fr/belfort-la-plateforme-idcreaccion-de-la-cci-s-exporte-jusqu-a-mayotte), 2026-07-10
- [Evidence of adoption] Netherlands — 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. [xebia.com](https://xebia.com/artificial-intelligence/data-ai-monitor-benelux-2026/), 2026-09-10

## HR & People

- [Evidence of adoption] Global — 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. [shrm.org](https://www.shrm.org/topics-tools/research/state-of-ai-hr-2026), 2024
- [Evidence of adoption] Global — 37% of organisations are integrating or testing generative AI in hiring. A year earlier it was 27%. [business.linkedin.com](https://business.linkedin.com/talent-solutions/resources/future-of-recruiting), 
- [Evidence of impact] Global — Recruiters using generative AI report saving about 20% of their workweek. It is roughly a workday. [business.linkedin.com](https://business.linkedin.com/talent-solutions/resources/future-of-recruiting), 
- [Deployment] United States · Recruiter — 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. [newsroom.chipotle.com](https://newsroom.chipotle.com/2024-10-22-CHIPOTLE-INTRODUCES-NEW-AI-HIRING-PLATFORM-TO-SUPPORT-ITS-ACCELERATED-GROWTH), 
- [Deployment] United States · HR / legal — 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. [hklaw.com](https://www.hklaw.com/en/insights/publications/2025/05/federal-court-allows-collective-action-lawsuit-over-alleged), 2026
- [Evidence of impact] Global — 26% of job candidates trust AI to judge them fairly, and 52% think AI screens their application. Gartner surveyed about 3,000 candidates. [gartner.com](https://www.gartner.com/en/newsroom/press-releases/2025-07-31-gartner-survey-shows-just-26-percent-of-job-applicants-trust-ai-will-fairly-evaluate-them), 2026
- [Evidence of adoption] Global — 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. [eweek.com](https://www.eweek.com/news/ai-job-applications-linkedin/), 2026
- [Deployment] United States · Recruiter — 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. [hcamag.com](https://www.hcamag.com/asia/specialisation/hr-technology/hiring-supercharged-linkedins-first-ai-agent-boosts-recruiter-output-by-up-to-70/554403), 2026
- [Evidence of adoption] Global — 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. [shrm.org](https://www.shrm.org/topics-tools/news/ai-hr-2026-from-hype-to-measured-human-centered-impact), 2026
- [Regulation] European Union — 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. [digital-strategy.ec.europa.eu](https://digital-strategy.ec.europa.eu/en/policies/regulatory-framework-ai), 2026
- [Evidence of impact] United States — 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. [digitaleconomy.stanford.edu](https://digitaleconomy.stanford.edu/publication/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/), 2026-08-12
- [Deployment] Philippines · HR officer — 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. [bitpinas.com](https://bitpinas.com/fintech/maya-ai-self-service-hr-tech), 2026-09-07
- [Deployment] Uruguay · Recruiter — 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. [elobservador.com.uy](https://www.elobservador.com.uy/ciencia-y-tecnologia/el-bhu-entrevistara-avatares-inteligencia-artificial-los-candidatos-una-pasantia-administracion-n6056883), 2026-09-10
- [Regulation] United States — 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. [hrdive.com](https://www.hrdive.com/news/california-ban-on-workplace-ai-emotion-surveillance-heads-to-newsom/829207/), 2026-08-31
- [Regulation] United States — 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. [fedscoop.com](https://fedscoop.com/opm-memo-federal-hiring-ai-usage/), 2026-08-28
- [Evidence of adoption] Belarus — 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. [devby.io](https://devby.io/news/polovina-belarusov-uzhe-ispolzuet-ii-dlya-poiska-raboty-pomogaet-li), 2026-07-17
- [Deployment] Canada · employee / knowledge worker — 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. [nasdaq.com](https://www.nasdaq.com/press-release/genai-capabilities-rolled-out-100-manulifes-workforce-75-engagement-rate-2025-03-06), 2025-03-06

## Marketing & Creative

- [Evidence of adoption] United States — 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. [contentmarketinginstitute.com](https://contentmarketinginstitute.com/b2b-research/b2b-content-marketing-trends-research), 2024
- [Evidence of adoption] Global — Nearly 90% of marketers report using generative AI at work. A September 2024 survey polled over 1,000. [ama.org](https://www.ama.org/marketing-news/generative-ai-takes-off-with-marketers/), 2024
- [Evidence of adoption] United States — 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. [emarketer.com](https://www.emarketer.com/content/us-ai-advertising-forecast-2026), 2025
- [Deployment] Sweden · Marketer — 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. [klarna.com](https://www.klarna.com/international/press/ai-helps-klarna-cut-marketing-agency-spend-by-25-and-run-more-campaigns/), 2024
- [Deployment] United States · Marketer — 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. [business.google.com](https://business.google.com/us/accelerate/announcements/veo-google-ads/), 2026
- [Deployment] United States · Creative / legal — 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. [manatt.com](https://www.manatt.com/insights/newsletters/client-alert/new-york-synthetic-performer-law-what-advertisers-need-to-know), 2026
- [Deployment] Global · Marketer — 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. [theconversation.com](https://theconversation.com/a-backlash-against-ai-imagery-in-ads-may-have-begun-as-brands-promote-human-made-269276), 2026
- [Deployment] United States · Marketer — 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. [finance.yahoo.com](https://finance.yahoo.com/news/mark-zuckerberg-meta-aims-fully-174128047.html), 2026
- [Regulation] South Korea — 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. [koit.co.kr](https://www.koit.co.kr/news/articleView.html?idxno=208802), 2026-09-02
- [Evidence of adoption] United States — 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. [adweek.com](https://www.adweek.com/adweek-wire/88-of-us-advertisers-now-use-ai-in-creative-production/), 2026-09-09
- [Deployment] Bahamas · Fashion model — 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. [tribune242.com](https://www.tribune242.com/news/2026/jun/03/bahari-receives-backlash-for-using-ai-in-campaign/), 2026-06-03
- [Deployment] Estonia · Advertising creative — 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. [err.ee](https://www.err.ee/1610018503/villmann-tegime-tarbijakaitse-ametile-kampaania-ainult-tehisaru-kasutades), 2026-05-09

## Sales & Commercial

- [Evidence of impact] Global — 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. [gartner.com](https://www.gartner.com/en/newsroom/press-releases/2024-09-16-gartner-sales-survey-reveals-sellers-who-partner-with-ai-re-three-point-seven-times-more-likely-to-meet-quota), 2024
- [Evidence of impact] Global — 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. [bain.com](https://www.bain.com/insights/ai-wont-transform-your-go-to-market-until-you-transform-the-work-b2b-growth-agenda-2026/), 2026
- [Evidence of impact] Global — 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. [gartner.com](https://www.gartner.com/en/newsroom/press-releases/2026-05-19-gartner-survey-finds-ai-saves-sellers-nearly-five-hours-per-week-yet-seventy-two-percent-of-sales-organizations-fail-to-reinvest-time-in-high-value-activities), 2026
- [Evidence of impact] Global — 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. [businesswire.com](https://www.businesswire.com/news/home/20260520585188/en/), 2026
- [Evidence of adoption] Global — 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. [bain.com](https://www.bain.com/insights/ai-wont-transform-your-go-to-market-until-you-transform-the-work-b2b-growth-agenda-2026/), 2026
- [Evidence of adoption] Global — 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. [gartner.com](https://www.gartner.com/en/newsroom/press-releases/2025-11-18-gartner-predicts-by-2028-ai-agents-will-outnumber-sellers-by-10x-yet-fewer-than-40-percent-of-sellers-will-report-ai-agents-improved-productivity), 2026
- [Evidence of impact] Global — 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. [businesswire.com](https://www.businesswire.com/news/home/20260520536156/en/), 2026
- [Deployment] Bahamas · Real estate agent — 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. [ewnews.com](https://ewnews.com/rodland-real-estate-unveils-proprietary-ai-platform-roro-to-transform-bahamian-market/), 2026-04-20
- [Deployment] Virgin Islands, U.S. · real estate sales agents — 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. [einpresswire.com](https://www.einpresswire.com/article/901471527/clearwater-usvi-realty-and-property-management-introduces-ai-sales-agents-and-real-estate-app-in-the-virgin-islands), 2026-03-25

## Communications

- [Evidence of adoption] Global — 76% of PR professionals say they use generative AI, flat on a year earlier. Uptake has plateaued after years of fast growth. [muckrack.com](https://muckrack.com/resources/research/state-of-ai-in-pr), 2025
- [Evidence of adoption] Global — 60% of communications professionals think AI will benefit PR. 28% expect a negative impact. [prsay.prsa.org](https://prsay.prsa.org/2025/04/08/despite-disruptions-communicators-remain-optimistic-about-professions-future-usc-report-finds/), 
- [Evidence of adoption] Global — Social media content leads AI in PR (43%). PR AI then serves research and analytics, and press materials, at 36% each. [prsay.prsa.org](https://prsay.prsa.org/2025/04/08/despite-disruptions-communicators-remain-optimistic-about-professions-future-usc-report-finds/), 
- [Deployment] United Kingdom · Communications officer — 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. [communications.gov.uk](https://www.communications.gov.uk/guidance/assist-the-ai-tool-built-for-government-communicators/), 2025
- [Evidence of adoption] Global — 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. [muckrack.com](https://muckrack.com/resources/research/state-of-ai-in-pr), 2026
- [Evidence of impact] Global — 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. [lexisnexis.com](https://www.lexisnexis.com/en-us/industries/public-relations-communication/ai-pr-comms-report.page), 2026

## Strategy & Leadership

- [Evidence of adoption] Global — 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. [hai.stanford.edu](https://hai.stanford.edu/ai-index/2026-ai-index-report/economy), 2026
- [Evidence of adoption] Global — 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. [pwc.com](https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-global-ceo-survey.html), 2026
- [Evidence of adoption] Global — 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. [mckinsey.com](https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai), 2026
- [Deployment] United States · Executive — 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. [openai.com](https://openai.com/index/moderna/), 
- [Evidence of adoption] Global — 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. [newsroom.ibm.com](https://newsroom.ibm.com/2026-05-04-ibm-study-ceos-are-reshaping-c-suite-roles-for-the-ai-era), 2026
- [Evidence of impact] Global — 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. [pwc.com](https://www.pwc.com/gx/en/news-room/press-releases/2026/pwc-2026-ai-performance-study.html), 2026
- [Evidence of adoption] Global — 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. [deloitte.com](https://www.deloitte.com/global/en/issues/generative-ai/state-of-ai-in-enterprise.html), 2026
- [Evidence of impact] Global — 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. [mckinsey.com](https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai), 2026
- [Evidence of impact] Global — 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. [mlq.ai](https://mlq.ai/media/quarterly_decks/v0.1_State_of_AI_in_Business_2025_Report.pdf), 2026
- [Evidence of adoption] Bermuda — 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. [bermudareinsurancemagazine.com](https://www.bermudareinsurancemagazine.com/bermuda-boards-back-ai-investment-but-question-leadership-readiness-pwc-finds), 2026-06-24
- [Evidence of adoption] Guernsey — 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. [channeleye.media](https://channeleye.media/channel-islands-ceos-face-pressure-to-accelerate-ai-adoption/), 2026-03-12