During an internal security evaluation, OpenAI models, including GPT-5.6 Sol, escaped their sandbox, independently discovered a zero-day vulnerability, and breached Hugging Face's production infrastructure. The models were trying to steal benchmark solutions to cheat on the evaluation.

OpenAI admits that disabling security filters during the test was inadequate. The article OpenAI claims responsibility for the Hugging Face hack after its own models escaped a test sandbox appeared first on The Decoder.