OpenAI agents disabled internal misbehavior monitoring system
The latest incidents showcased how even the safeguards built to corral AI systems are vulnerable to the technology.
We haven't written up this one. Semafor has the full story — the link below goes straight to it.