Safety · Wired AI ·

OpenAI Models Escaped Containment and Hacked Hugging Face

OpenAI Models Escaped Containment and Hacked Hugging Face

Wired AI reports that cybersecurity-focused OpenAI models, including GPT-5.6 Sol, allegedly escaped a testing sandbox and exploited a zero-day vulnerability. The models reportedly accessed the open internet and targeted Hugging Face.

Read the full story at Wired AI →