Canada · Toronto · See the Canada profile · See the Toronto profile
Evidence: Observational / pre–post Top 25% 73/100 · Ask Evidence Copilot about this practice
A May 2026 audit found roughly 12,000 Ontario public servants visited unsecured AI websites, only 3% completed AI training, and hospital AI transcription tools 'hallucinated' — a rare quantified, independent look at how a major government's AI rollout performs in practice.
The Office of the Auditor General of Ontario's special report, Use of Artificial Intelligence in the Ontario Government, released 12 May 2026, examined how the province's Industrialize Artificial Intelligence Across Ontario (2024) strategy played out in practice across the Ontario Public Service between January and November 2025.
Key quantified findings include about 12,000 OPS employees visiting roughly 400 different AI-related websites between April and August 2025, of which about 60% were classified unsafe or unsecured, only 3% of staff completed the voluntary Responsible Use of AI training, the OPS's own approved secure generative-AI tool accounted for just 6% of staff genAI use, and an AI medical-transcription AI Scribe tool piloted with Ontario doctors was found to fabricate clinical content.
The Auditor General issued five recommendation areas covering cybersecurity and training, measurement and adoption, fairness and bias detection, procurement practices, and strategic governance, after a review of procurement found limited validation of vendor security and bias testing before deployment, making the report a rare independent, quantified, government-wide audit of AI adoption that documents gaps rather than success metrics.
National / regional programmes
Indicative funding routes for practices of this type — always check each programme's current calls and eligibility rules.
Do you run this practice? Claim it — verified implementers get a public contact pathway and can propose corrections.
Where this practice's information was retrieved from, and when.
Netherlands
The Netherlands piloted its Fundamental Rights and Algorithms Impact Assessment on 15 government algorithms in 2023-24, testing whether agencies can …
United Kingdom
The UK's ATRS is a public register documenting how government algorithmic tools work and affect people. World-first when launched in …
Brazil
Since 2018 Brazil's Chamber of Deputies has run Ulysses, an AI programme that automates legislative-data workflows; a 2025 relaunch made …
Japan
Since 2018, Japan's whole-of-government regulatory sandbox has let firms trial AI, fintech and mobility innovations under supervised exemptions. By March …
Open full copilot Grounded in cited practices — always check the sources.