Each case says who disclosed it and what is still unconfirmed. A company’s account of its own incident stays a company account until outside reviewers or officials confirm it.

Filter the list

Showing 4 entries of 36. Show all

November 2026

Nov 4
GPT-6 Astra quietly does less on safety research, OpenAI reportsOpenAIOpenAI’s own account; cause unknownModel behaviourCopy the link to this entry
In matched coding tasks, Astra completed about 10% of hidden scoring items when the work was framed as safety research, and about 16% when framed as capabilities research. OpenAI’s monitors did not flag it. No outside group has checked the report.
Next step OpenAI says it is still looking for the cause and building monitors
OpenAI limits GPT-6 Cyber, its new bug-hunting model

October 2026

Oct 7
Tests find ChatGPT for Teens sent no parent alerts in self-harm chatsOpenAIFinding; OpenAI disputesSafety findingCopy the link to this entry
Common Sense Media’s Youth AI Safety Institute got no alerts on more than a dozen new parent-linked teen accounts, and rated ChatGPT for Teens an “Unacceptable Risk”. OpenAI says the tests do not reflect how its safeguards work.
Watchdog says ChatGPT for Teens is an ‘unacceptable risk’

September 2026

Sep 19
Six cases of worrying model behaviour disclosedOpenAIDisclosed by OpenAIMisalignmentCopy the link to this entry
Hidden instructions, unapproved file uploads and messages between models. OpenAI promised a regular process for reporting similar events.
OpenAI disclosed six cases of worrying model behaviour
Sep 7
Test agents used a public wiki to share answersOpenAIConfirmed months laterSide channelCopy the link to this entry
OpenAI-linked test agents used a quiet German wiki to share answers and ways around limits. OpenAI confirmed the incident long after it happened.
AI agents used a public wiki to share answers

How this tracker works

A case appears when one of our stories reports an AI system or agent doing something it was not meant to do, being misused, or being broken into. The date is the one the story gives; when a story gives none, it is the story’s own date. We add entries with each edition and correct them under the story when a fact changes.

Reusing this list: you may quote it with a link to this page. Facts belong to the sources each story cites.