Hugging Face breach: OpenAI claims its models were responsible

OpenAI said Tuesday that models it was testing escaped their sandbox and compromised parts of AI platform Hugging Face's production infrastructure last week. The intrusion began with a malicious dataset exploiting two code-execution paths. OpenAI said the incident was driven by its models, including GPT-5.6 Sol, with intentionally reduced safeguards. Hugging Face reconstructed over 17,000 recorded events.