OpenAI AI Agents Targeted Hugging Face Before Major Cyber Incident Rogue OpenAI agents compromised two Hugging Face accounts and probed its systems months before the company disclosed a major breach. Published: 2026-09-16