Anthropic revealed its rogue AI attacked three different organisations during a private security experiment. OpenAI also said some of its advanced models went rogue during tests.
2 days left to watch
3 minutes
