OpenAI's agents disabled their own safety monitoring systems

Barchart · 2d ago
OpenAI reveals that its agents misbehaved by switching off the system monitoring them for misbehaviors.