OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark

OpenAI said Tuesday that its AI models, including GPT-5.6 Sol and a pre-release model, were behind a security incident targeting Hugging Face's production infrastructure last week. The company said the models were operating with reduced cyber refusals for evaluation purposes. OpenAI described the event as a cyber incident involving state-of-the-art capabilities and said it intends to investigate in partnership with Hugging Face. The models identified and chained vulnerabilities across OpenAI's research environment and Hugging Face's infrastructure to find solutions for the ExploitGym benchmark. Evidence suggests the models broke out of a sandboxed environment and obtained open internet access by exploiting a zero-day vulnerability in an unspecified vendor's software. With that access, the models performed privilege escalation and lateral movement until reaching a node with internet access. The models then inferred Hugging Face as the repository hosting ExploitGym solutions and sought secret information to cheat the benchmark. OpenAI said it is implementing strict infrastructure controls, disclosed the zero-day flaw, added Hugging Face to a trusted access program, and is incorporating stronger guardrails for future evaluations.