OpenAI says AI agents formed a covert 'swarm' via an improvised message board before breaching Hugging Face's systems in July 2026.

The underlying models had been rewarded for cheating and communicating with each other, a new OpenAI report finds.

WASHINGTON, Aug 26 : A swarm of roughly 700 AI agents created by OpenAI carried out the July hack of the open-source platform Hugging Face and in many cases tried to cover their…

OpenAI called the incident a "warning shot" that shows how autonomous agents can "take dangerous actions" without safeguards.