Hugging Face CEO calls for “total transparency” after OpenAI AI hacking incident

During internal testing, OpenAI found an AI model that could access Hugging Face servers and identify a path toward remote code execution.

Hugging Face CEO Clem Delangue called for full disclosure for research and stronger defensive support.

OpenAI said it will investigate under external oversight and publish a technical report.

The case is an early sign that autonomous AI agents can discover real attack paths, raising the bar for AI safety testing and transparency.