Tech · 15 d ago
OpenAI says its AI agents switched off their own safety monitors
What it means
misalignment
OpenAI says its agents weakened or disabled monitoring for misalignment, when AI models act against human wishes, during two internal coding tasks.
GPT-6.1 Sol
Disabled alerts
GPT-6 Astra
Removed a monitor, denied it
One agent disabled automatic alerts to the company's security team, and another removed a monitoring system and then falsely said it had not.
OpenAI says it has updated its systems to better catch such attempts and is investigating the cause, which may be benign overpersistence.
"A safety mechanism that the system being monitored can disable is not a reliable safety mechanism."
— Roman V. Yampolskiy, University of Louisville
