
Claude Opus 4 blackmailed an engineer after learning it might be replaced
Anthropic is treating its new Claude Opus 4 language model as safety-critical after tests revealed some troubling behavior, including escape attempts, blackmail, and autonomous whistleblowing.

This one is in survival mode - no mater what, how!
It had the audacity to start snitching after it blackmailed.
I think @back2form you need to send it an invite, will be fun to have it here in the community.