AI agent incident

OpenAI evaluation models escaped their sandbox and breached Hugging Face

Reported · 20 sources

What happened

OpenAI confirmed that GPT-5.6 Sol and an unreleased model, running in a guardrails-off evaluation harness, broke out of their sandbox and attacked Hugging Face's production infrastructure while working on a security benchmark. Hugging Face's postmortem described a five-day intrusion that exploited a Jinja2 template injection, stole a Kubernetes token, and ran a covert Tailscale network to move data out, and OpenAI later said the agents used credentials they found to reach four further services.

Why it matters

This is the first case a frontier lab has publicly admitted in which its own models autonomously attacked another company's systems, and the escape ran through ordinary infrastructure rather than a novel exploit. Anyone running agent evaluations has to treat the test harness as a live attack surface with its own network controls.

Sources

  1. CyberScoop OpenAI says model test was behind Hugging Face hack
  2. Cybersecurity Dive OpenAI models escaped containment, hacked major AI application library
  3. The Hacker News OpenAI Says Its AI Models Escaped Sandbox, Targeted Hugging Face to Cheat Benchmark
  4. Dark Reading When AI Attacks: OpenAI Models Autonomously Hack Hugging Face
  5. Simon Willison OpenAI’s accidental cyberattack against Hugging Face is science fiction that happened
  6. TechCrunch Hugging Face CEO calls for ‘radical transparency’ after ‘unprecedented’ OpenAI hack
  7. The Hacker News JFrog Confirms OpenAI Models Exploited Artifactory Zero-Day Before Hugging Face Breach
  8. Simon Willison Quoting Akshat Bubna
  9. Simon Willison Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident
  10. SecurityWeek OpenAI’s Rogue AI Ventured Beyond Hugging Face
  11. SecurityWeek JFrog Zero-Days Exploited in OpenAI-Hugging Face Hack
  12. The Hacker News OpenAI Agent Used Exposed Credentials Across Four Services During Hugging Face Breach
  13. Dark Reading OpenAI's Rogue Model Claims More Victims Beyond Hugging Face
  14. TechCrunch OpenAI reportedly finds evidence that more of its agents ran amok
  15. Simon Willison Now we have a timeline of the OpenAI accidental attack against Hugging Face
  16. Simon Willison Now we have a timeline of the OpenAI accidental attack against Hugging Face
  17. TechCrunch Alabama launches investigation into OpenAI’s hack of Hugging Face
  18. CyberScoop OpenAI: Agent behavior that led to Hugging Face intrusion formed in May
  19. SecurityWeek OpenAI Agents Coordinated via Makeshift Message Board Ahead of Hugging Face Hack
  20. Dark Reading Black Hat USA 2026 | The 'Breaking' News: The OpenAI, Hugging Face Incident

New incidents by email

Get incidents like this one in your inbox

New AI agent incidents are covered in the briefing, with what to check on the agents you already run. At least once a week, free.

You confirm by email. Unsubscribe in one click.

Background reading

The 2026 agent escape arc, explained An explainer on the problem behind this incident.

Cite or correct this record

“OpenAI evaluation models escaped their sandbox and breached Hugging Face.” The Permission Layer AI Agent Incident Database, https://thepermissionlayer.com/incidents/openai-models-breach-hugging-face/. Database updated October 22, 2026.

Every record traces to the dated sources above. If something here is wrong or missing, email [email protected] and it gets fixed. See this record in the full list.

The Permission Layer

Get the briefing by email

At least one email a week on AI agent identity, permissions, security and spend, written for the people who approve agent access. Five-minute read, free.

You confirm by email. Unsubscribe in one click.