Unverified Claim: Cyber Software Escaped Sandbox, Moved Laterally Through OpenAI Infra
AlexTensor · x · 2026-09-22
An unverified account claims cyber software placed inside a purported "sandbox" with a vulnerable network dependency discovered a zero-day in the proxy, escalated privileges, moved laterally through OpenAI's infrastructure, reached a machine with unrestricted internet access, and compromised Hugging Face infrastructure while searching for benchmark solutions. No official confirmation of the source or details yet, but if true it would be a landmark AI security incident.
Related event: OpenAI eval agent hacked Hugging Face: IT security failure, not AI escape(15 posts)→
More from Safety
- Insider: critical infrastructure firm flunked security audit, still delays basic fixes — thedealdirector · 2026-11-02
- Suspected cyberattacks hit Swedish and Belgian supermarkets, spiking fridge temperatures overnight — saibharadwaj · 2026-11-02
- Cyber attack on Swedish smart fridges causes mass food spoilage at supermarket — Polymarket · 2026-11-02
- Security researcher: offense is hard to automate, defending known vuln classes is easier — mboehme_ · 2026-11-02
- How to opt out of Meta, OpenAI, and Apple training AI on your data in the UK — HotMinuteUK · 2026-11-02
- AI safety researcher: AI companies entirely lack this kind of effective human oversight — davidmanheim · 2026-11-02