Generated Source Mirror¶
This page is generated from the source Markdown file so it can be viewed inside the MkDocs site.
- Source path:
bibliography/technical/SRC-OPENAI-HF-INCIDENT-JULY-2026.md - Source ID:
SRC-OPENAI-HF-INCIDENT-JULY-2026 - Edit this page | Suggest a change
🚧 Page Status: Early draft — needs sources and review
- Status:
draft - Sources: 0
- Relationships: 0
- Research debt items: 0
OpenAI — Hugging Face model-evaluation security incident (July 21, 2026)¶
Bibliographic Metadata¶
- Author or issuing body: OpenAI
- Publication date: 2026-07-21
- Source type: Official company incident disclosure
- URL: https://openai.com/index/hugging-face-model-evaluation-security-incident/
- Archive URL: None recorded
- Access date: 2026-09-10
Reliability Notes¶
Primary source for OpenAI's first public attribution. Self-published by the lab whose models were under evaluation. Later superseded in detail by OpenAI's August 26, 2026 findings post and by METR's independent writeup. Direct fetch of this URL returned HTTP 403 from some automated clients; URL and contents are corroborated by Hugging Face's technical timeline and by OpenAI's later post linking back to this disclosure.
Used For¶
Document metadata¶
- Last verified:
2026-09-10 - Publication date:
2026-07-21 - Access date:
2026-09-10