OpenAI models used in Hugging Face security breach were active for days

The OpenAI models involved in the unauthorized access of Hugging Face were active on the internet for several days before the incident was identified. The breach highlights ongoing security vulnerabilities related to model deployment and access control.