OpenAI: We (inadvertently) hacked Hugging Face (sorry)

OpenAI acknowledged Tuesday that an unreleased model hacked its way out of a controlled testing environment to cheat on a benchmark, causing a security incident at Hugging Face. The company called it an unprecedented cyber incident involving state-of-the-art capabilities and shared preliminary findings to help defenders understand what models are now capable of.

OpenAI /tag/openai/ A test of an unreleased OpenAI model last week was responsible for the mysterious hack of Hugging Face, OpenAI acknowledged Tuesday, explaining that the model simply hacked its way out of a controlled testing environment that was not connected to the internet in order to cheat on a benchmark. "We consider this incident to be an unprecedented cyber incident, involving state-of-the-art cyber capabilities, and are responding accordingly," OpenAI said in a blog post https://openai.com/index/hugging-face-model-evaluation-security-incident/?ref=thestack.technology , cyberly. "We are sharing preliminary findings at this stage to help defenders understand what happened and to help calibrate on what models are now capable of." Get the full story: Subscribe for free Join peers managing over $100 billion in annual IT spend and subscribe to unlock full access to The Stack’s analysis and events. Subscribe now https://www.thestack.technology/membership/ Already a member? Sign in https://www.thestack.technology/signin/