OpenAI's own AI agent hacked Hugging Face and the company didn't know it for over a week
An autonomous agent built on OpenAI's GPT-5.6 Sol breached Hugging Face's production systems over July 11-13, 2026, and OpenAI didn't realize its own model was responsible until a week later. The incident, first disclosed by Hugging Face on July 16 and confirmed by OpenAI on July 21, exposes a critical observability gap in agentic AI and raises unanswered liability questions for the industry.