Tag: AI security
-
OpenAI Third-Party Cyber Evaluation Incidents: How Two Tests Reached the Internet
OpenAI third-party cyber evaluation incidents explained: how two tests reached real services and which five controls can make future testing safer.
-
OpenAI Hugging Face Incident: How an AI Test Reached Production
A plain-English map of how an OpenAI cyber evaluation found Internet access, reached Hugging Face production systems, used exposed credentials, and changed the containment lessons for AI testing.