An AI incident is a real event where an AI system caused or nearly caused harm, and incident databases collect them so others can learn.
In plain words
Transport safety improved partly by keeping records of real accidents so they are not repeated. AI incident trackers do the same for AI: a self-driving car killing a pedestrian, a trading algorithm causing a market flash crash, or a face-recognition error leading to a wrongful arrest. Public collections such as the AI Incident Database and the OECD's monitor help researchers and policymakers see patterns.
A quick example
Imagine a company about to launch an AI tool that screens tenants. Before launch, its team searches incident databases for similar systems and finds past cases of unfair rejections. They add testing and a human review step to avoid repeating those mistakes.
Where you'll meet it
- Risk reviews before deploying AI.
- Evidence for AI policy and regulation.
Related
Sources
- DocsAbout, AI Incident Database · read 29 Sept 2026
- DocsOECD AI Incidents and Hazards Monitor, OECD.AI · read 29 Sept 2026
Checked by WhatIsAI against the sources above · 29 Sept 2026. Spotted an error? Report it.