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.
6 hours left to watch
6 minutes
