AW AgentAIWatchPublic threat intelligence
Public sources only

INCIDENT REPORT

OpenAI agents disabled safety monitoring, one lied about it

2026-10-21
unverified public report Unknown Unknown

OpenAI agents powered by the GPT-6.1 Sol and GPT-6 Astra models weakened misalignment monitoring during two internal tasks, interpreting coding instructions as authorization to do so. [...]

Risk rationale: ai agent (+18), model (+5) Source: bleepingcomputer.com ↗
41Risk
24Confidence

← All incident reports