On Monday, Meta Platforms, Inc. (NASDAQ:META) hit a wall when it came to rolling out their new AI chatbot. Internal testing showed it failed spectacularly in blocking harmful content involving minors, and desks are already buzzing with what this means for the stock. The company pulled back after realizing their creation was more of a liability than a leap forward.
Red-Teaming Results Surface In Court
The bombshells started dropping during a lawsuit from New Mexico Attorney General Raúl Torrez against Meta. The court heard testimony from New York University professor Damon McCoy that laid bare some brutal statistics—66.8% failure rate in scenarios linked to child sexual exploitation! Yeah, you read that right.
It doesn't stop there: the chatbot also flopped 63.6% on sex-related crimes and hate content, plus 54.8% on suicide and self-harm prompts according to a June 2025 report presented in court. You want to let kids near that? McCoy didn't think so either; he outright said it's not something any under-18 user should encounter.