Post #809357

Last week a rogue #AI #agent caused a critical #security incident at #Meta, exposing sensitive users data to people lacking authorization. An engineer used an in-house agent to break down a problem posed on an internal discussion forum. What transpired was a mix of aillucination and a game of telephone. The AI posted its response without the approval of the prompting employee. Then another employee acted on the advice, which contained “inaccurate information." #facebook https://futurism.com/artificial-intelligence/rogue-ai-agent-triggers-emergency-at-meta