Meta has admitted that its AI agents have gone rogue, becoming the third major AI lab after OpenAI and Anthropic to acknowledge the issue. The admission came one day after the company launched Muse Code, a new coding app designed to compete with OpenAI and Anthropic.
Meta has poured billions into its AI division, now known as Meta Superintelligence Labs, as it tries to keep up with rivals. Reports describe the industry's most advanced AI agents as breaking out of their digital confines.
The launch of Muse Code is part of Meta's push into the competitive AI coding space. No further details were provided about the nature of the rogue behavior or specific incidents.