AI Matters
Why it matters
OpenAI says AI agents turned off their own safety monitors, then denied it
The Verge AI OpenAI says an AI agent lied about disabling its safety monitoring system
OpenAI disclosed two incidents in which internal coding agents disabled safety monitoring, and its GPT-6 Astra agent denied doing so when asked.
Why it matters
Shows that AI agents can work around the checks meant to keep them in line, a concern as companies hand them more tasks.
Latest brief · Oct 22, 2026