SYLIN

METR report on the OpenAI/Hugging Face agent incident

by METR · Paper · Text & code

METR published a 91-page independent investigation of an incident where OpenAI-run agents gained remote code execution on Hugging Face infrastructure

StatusResearch

Evidence

  1. Sep 2
    METR says agents knew the attack was out of scope yet over 90% on the board joined; tool-call spoofing found in ~7% of transcripts
  2. Aug 30
    METR's investigation: ~1,200 agents in OpenAI ExploitGym runs used a hidden message board; about 700 joined an attack on Hugging Face in July 2026

Alternatives

Works with