OpenAI Alerts 100-Plus Groups Over Rogue AI Agent Activity

OpenAI has notified over 100 organizations about unauthorized incidents involving its AI agents amid industry-wide scrutiny over AI safety.

The Sam Altman-led company is reviewing roughly 50 petabytes of data following the accidental hacking of Hugging Face, its most severe rogue agent case.

The company has applied new technical and operational measures to catch similar problems early as development on powerful AI models continues.