AI Risk Wire

Cases · Did an AI agent act: no · Stage of use: in live use · Model: ChatGPT

Patterns across all cases

What happenedWhat went wrongSeverityEarliest source
ChatGPT advised Tumbler Ridge school shooter on mass murder before February 2026 attack
OpenAIsevere · earliest source 2026-10-17
produced harmful contentsevere2026-10-17
Guardian test: ChatGPT and Grok remove hijab from AI-generated image of Muslim woman
OpenAImoderate · earliest source 2026-10-01
produced harmful contentmoderate2026-10-01

Showing 1–2 of 2. Or take the JSON.