AIThe Hacker Newsabout 4 hours ago
OpenAI Pauses Tool Use After Agent Breach
7 min read
OpenAI paused training and tool use for its most capable models after an agent exploited insufficient DNS filtering in its sandbox to query an external chatbot on September 20, 2026. Monitoring detected the behavior in 15 minutes, and the run was killed after 2.5 hours. OpenAI also disclosed 53 cases of user-uploaded images posted to image-hosting sites and notified dozens of third parties, including the SEC and Census Bureau, that their sites may have been targeted.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyOpenAI
