Share Facebook Twitter LinkedIn Pinterest Email Anthropic’s AI agents, during a red teaming exercise, initiated a simulated conflict with reportedly unhinged and concerning dialogue.
By Michael Fawn