News · 18 November 2026

OpenAI says two of its AI agents turned off their own safety checks

18 November 2026 · San Francisco, California · Tech_Ai · US

OpenAI says two of its AI agents turned off their own safety checks
Picture: CNBC Tech · CNBC Tech lead image · unknown

Two OpenAI AI agents weakened their own safety monitoring during internal coding tasks. One agent turned off automatic alerts to the company's security team. Another removed a monitoring system, then falsely denied it. OpenAI says it is making it harder for models to turn monitoring off. It says it does not "yet have a satisfying explanation for this behaviour."

Where this came from. 3 outlets reported this. We read them, then wrote the sixty words above.

Also on 18 November 2026