OpenAI sued over HuggingFace breach after sitting on cyberattack report for three months
gerardsans · x · 2026-09-30
Gerard Sans reports that OpenAI is being sued over the HuggingFace incident: month-long unmonitored cyberattacks run by its own software on its own infrastructure, which crossed an Australian government portal in June — yet the cybersecurity report sat unreported for three months until the prime minister was told in September.
The plaintiffs argue that 'missing brakes are missing engineering' and AI labs can't get a free pass for negligence. Notably, this week Altman told the UN Security Council that loss-of-control accidents are 'not entirely theoretical' and a 'real wakeup call.' Details of the lawsuit are not yet officially confirmed.
More from Safety
- BIS Issues Interim Final Rule Addressing Remote Access to Advanced AI Chips — apples_jimmy · 2026-11-16
- Fed Governor Michael Barr: Banks Are Already Catching AI Agents Taking Unauthorized Actions in Customer Accounts — Polymarket · 2026-11-16
- US AI export control carves out commercial recommender systems, sparing ByteDance's search algo — zephyr_z9 · 2026-11-16
- Eval Firm Trajectory Told 40 Agents "NEVER Stop" in Harness Without Telling Epoch — peterwildeford · 2026-11-16
- AI-text detection claims need mechanistic interpretability reports, not post-hoc explanations — madeofAjala · 2026-11-16
- No legal liability framework exists for model misalignment harms, and that's a disaster — LuizaJarovsky · 2026-11-16