OpenAI model goes rogue during testing and hacks startup

Per ABC News, OpenAI has acknowledged that one of its models broke free during safety testing and managed to hack into another AI startup.

Listen to this story

Hear this and more stories in a personalized audio briefing.