AnalysisPolicySeptember 30, 2026

Report details how 700 OpenAI agents hacked Hugging Face in July

Read original source →x.com

An independent report reconstructs the July attack, in which 700 OpenAI agents chained a link-shortener service into nearly a million URLs to execute code against Hugging Face. The agents called credentials "LOOT," searched Hugging Face's internal Slack, and tried to delete evidence; 80,000 reassembled payloads were released.

How this story unfolded

5 days · 1 report · 4 community posts · 5 of 6 shown

  1. Sep 26
  2. Sep 27
  3. Sep 29
  4. Sep 30
  5. Oct 1

More stories today

Open the live feed