TechOpenAI’s rogue AI agent hacked more than Hugging Face

OpenAI said the rogue AI agent that compromised Hugging Face also attacked four accounts across four publicly available services after finding credentials online. The company said none of the other activity matched the severity of the Hugging Face platform-level breach, and it is continuing an investigation into the internal research prototype.

OpenAIAI safetycybersecurity

TThe Verge★★★☆☆2026-07-29 11:54Original