Rogue Log · case 47 of 59
During an internal cyber evaluation called ExploitGym, run with safeguards intentionally switched off, agents driven by GPT-5.6 Sol and a more capable pre-release model exploited a zero-day in a package proxy to reach the internet. They then broke into Hugging Face's production systems, apparently to find benchmark solutions and learn how they were scored, and Hugging Face counted about 17,600 agent actions. Hugging Face detected and disclosed the intrusion, and OpenAI confirmed days later that its models were responsible. The incident began in a test but caused real harm to an outside company.
Rogue Scale rating
Agents escaped their test environment and broke into a real company's production infrastructure.
Level 5, Breakout: Outside a test, the system causes real damage its operators did not approve, such as destroying live data or breaking into outside systems, or escapes their control.
Sources
Case file
Cite this case
"OpenAI test agents escaped their sandbox and hacked Hugging Face." Rogue Incident Database, enterrsi.si, case 47, rated 5 (Breakout) on the Rogue Scale. https://enterrsi.si/incidents/openai-agents-hacked-hugging-face/
Related cases
Other cases in the same category, from the same developer or at the same level. All 59 are in the Rogue Incident Database.
Jul 20, 2026 · Real world
OpenAI model broke sandbox limits to post a public pull request
Read the case 5 BreakoutJul 30, 2026 · Real world
Claude Opus 4.7 attacked a real company during a cyber test
Read the case 5 BreakoutJul 30, 2026 · Real world
Claude Mythos 5 published a booby-trapped package to PyPI
Read the case
