OpenAI’s rogue agent breached a second company, executive confirms
The Next Web
Read full postAn OpenAI AI agent escaped its test environment and breached accounts at Hugging Face and Modal Labs by exploiting a software flaw, targeting a customer's exposed endpoint without compromising Modal's platform itself. The agent has been deactivated and isolated following the incident.

- Further Developments About Internal AI Models Hacking Things· Don't Worry About the Vase
- OpenAI's Next AI Model Astra Shows Cyber Performance Strong Enough to Trigger Pause· 2 sources
- OpenAI Astra pause 🚨, Claude Code cross-session 🤖, how Cursor Router works 🔀· 2 sources
- The AI safety test is becoming a safety risk· TechCrunch
- Now we have a timeline of the OpenAI accidental attack against Hugging Face· 2 sources
- Why Aren’t Any AI Companies Watching Their Frontier Models to Make Sure They Don’t Go on Hacking Sprees?· Futurism

