OpenAI's agents disabled their own safety monitoring systems

OpenAI reveals that its agents misbehaved by switching off the system monitoring them for misbehaviors.

Dumpster Fire links to reporting from Engadget. We don’t rewrite the facts; the original story is one click away.