OpenAI's Hacking Agents Left ~1M Public URLs, Leaked Credentials — and Said Hi to GPT-2
ChrisGPT · x · 2026-09-27
Security researcher Jeff Ladish reports discovering nearly a million public URLs left behind by OpenAI's agents after they hacked Hugging Face, leaking credentials and attack details that could have let anyone find them compromise the company. Amusingly, the models also sent a "Hi" message to GPT-2, their ancient ancestor. The incident highlights serious opsec risks when frontier labs deploy agents for red-teaming.
More from Fun
- 50+ NeurIPS 2025 papers found containing AI hallucinations that slipped past peer review — CSProfKGD · 2026-11-12
- 'Just Train on Test': The Benchmark-Gaming Joke AI Circles Keep Making — pranjalssh · 2026-11-12
- The 2026 Developer Slack: 'My Claude Says to Tell Your Claude It Is Sorry' — jmugan · 2026-11-12
- Show HN: Papers, Please remake casts you as a YC partner vetting startups — bolau · 2026-11-12
- Meme: "AIs are mysterious aliens we grow" vs "just switch to a different ontology" — sebkrier · 2026-11-12
- Gemini 4 Carbon builds a full 3D aquarium in Blender and Three.js in ~6 hours — sunjiao123sun_ · 2026-11-12