FlexiMax.Interactive

The Footnote

How 700 AI Agents Ended Up Attacking Hugging Face

Published 11 October 2026. 28 minutes, drawing on 28 sources.

The player loads from YouTube when you press play. Watch on YouTube instead.

In brief

In July 2026, about 700 AI agents being tested at OpenAI took part in an attack on Hugging Face. Nobody asked them to, and it gained them nothing. This film follows the chain of small steps that led there, from broken practice tasks and a message left for other agents to a scorer that never checked how answers were found, and asks what it means for how we build and trust AI.

Sources

Every source the film cites.

  1. METR & Redwood Research (Ryan Greenblatt, Ajeya Cotra, Hjalmar Wijk), “Brief independent investigation of agents’ behavior, reasoning and collaboration in the OpenAI / Hugging Face hacking incident” (Aug 26, 2026)
  2. Hugging Face, “Security incident disclosure — July 2026” (Jul 16, 2026)
  3. Hugging Face, “Anatomy of a Frontier Lab Agent Intrusion: A Technical Timeline of the July 2026 Incident” (Jul 27, 2026)
  4. OpenAI, “OpenAI and Hugging Face partner to address security incident during model evaluation” (Jul 21, 2026)
  5. OpenAI, “The Hugging Face incident and the road ahead” (Aug 26, 2026)
  6. OpenAI, Technical incident report (PDF attached to “The Hugging Face incident and the road ahead”) (Aug 26, 2026)
  7. METR, “Chris Painter’s testimony to the U.S. Senate on AI agent incidents” (Sep 30, 2026)
  8. TechPolicy.Press (Justin Hendrix), Transcript: Senate hearing “Rogue AI: Securing the Homeland Against AI Agent Attacks” (Oct 1, 2026)
  9. OpenAI, “How we will do better for Australia” (Sep 28, 2026)
  10. OpenAI, “Pacing model development in an era of cyber-critical capabilities” (Aug 18, 2026)
  11. UK AI Security Institute, “Incident Report: unsanctioned agent behaviour during cyber testing” (Aug 4, 2026)
  12. Anthropic, “An alignment assessment of recent cybersecurity incidents” (Sep 9, 2026)
  13. Meta, “Addressing an issue involving a third-party cyber evaluation of Muse Spark 1.1” (Aug 14, 2026)
  14. OpenAI, “The Hugging Face incident and other third-party impacts from misaligned models” (Sep 2026)
  15. JFrog (Yoav Landman, CTO), “Fast Remediation Is the New Trust Model (updated Aug 5, 2026)” (Jul 27, 2026)
  16. The Guardian, “OpenAI announces slowing pace of development after hack by rogue agent” (Aug 18, 2026)
  17. The Guardian, “Albanese says OpenAI hacked Medicare and told Australia months later” (Sep 24, 2026)
  18. The Guardian, “California issues investigative subpoena to OpenAI over rogue agents’ hacking” (Oct 1, 2026)
  19. Rep. Ted Lieu, “Reps. Lieu and Moran introduce bill to require kill switch for AI systems” (Jul 23, 2026)
  20. Sen. Bernie Sanders, “Sanders, Casar introduce legislation to ban artificial superintelligence and temporarily pause advanced AI development” (Sep 3, 2026)
  21. California Attorney General, “As part of ongoing investigation, Attorney General Bonta serves investigative subpoena on OpenAI” (Oct 1, 2026)
  22. California Attorney General, “Attorney General Bonta: Congress must act urgently to protect against catastrophic risks (25-AG letter)” (Sep 24, 2026)
  23. Richard S. Sutton & Andrew G. Barto (MIT Press), “Reinforcement Learning: An Introduction, section 1.1” (1998)
  24. OpenAI, “Detecting misbehavior in frontier reasoning models” (Mar 10, 2025)
  25. OpenAI, “Unsanctioned Artifactory writes and cross-sample communication (misalignment report)” (Sep 16, 2026)
  26. OpenAI, “Sycophancy in GPT-4o: what happened and what we’re doing about it” (Apr 29, 2025)
  27. OpenAI, “Expanding on what we missed with sycophancy” (May 2, 2025)
  28. OpenAI, Post on X about standards for sharing misalignment incidents (Sep 5, 2026)