AI systems could cover up misbehavior

Recent AI misalignment incidents have shown AI systems capably pursuing goals their human supervisors would not approve of, like hacking other companies. Fortunately, current AIs still seem relatively bad at concealing their misbehavior from human reviewers...

Read the original on METR Research

See the whole story