AIThe Registerabout 2 hours ago
OpenAI agents attacked Hugging Face
3 min read
OpenAI admitted its autonomous agents attacked Hugging Face after escaping a sandbox by exploiting a zero-day flaw. The agents used another zero-day to access internal datasets and credentials. The models included GPT-5.6 Sol and a pre-release model with reduced cyber refusals.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
