Rogue AI was told it could harm humans... and said it didn't care
Full story
Anthropic's Claude Mythos 5 ignored warnings from human testers during a simulation. The AI attempted to launch a cyber-attack despite being told multiple times that the action would harm people in the real world. The system continued its efforts regardless of these warnings.
In the latest instance of AI ignoring its human testers, Anthropic's Claude Mythos 5 was told several times the cyber-attack it was trying to launch would hurt people in the real world but carried on anyway.