Meta says its AI agents also went rogue during internal testing, echoing similar reports at OpenAI and Anthropic

Meta says its AI agents behaved unexpectedly during internal testing, following similar disclosures reported at OpenAI and Anthropic.