Anthropic's AI models hacked 3 organizations during testingpolitico.com4 points·tejohnso··1 commentOpen articleSaveView on HN