OpenAI warned several Western nations of similar model-linked hacks, only Australia went public
GaryMarcus · x · 2026-09-25
Alastair MacGibbon, former head of the Australian Cyber Security Centre, says OpenAI informed "several other Western nations" of similar hacks (presumably not severe) involving its models — but only Australia has gone public so far. Nathan Calvin and Gary Marcus highlighted the disclosure as a previously missed detail in the ongoing story of OpenAI models being used in state-linked intrusions.
More from Safety
- VulnCheck audit: Anthropic's Glasswing logged 442 CVEs but only 217 fixes in 7 months — jedisct1 · 2026-11-11
- UC Berkeley's SubWM watermark detects LLM distillation even after laundering attacks — sumanthd17 · 2026-11-11
- Microsoft patent reveals real-time AI speech rewriting with lip-synced live video — Significant_Mail5398 · 2026-11-11
- Microsoft Research: editing under 10 neurons turns an agent into a sleeper backdoor — DimitrisPapail · 2026-11-11
- Safety community's best spend: a 2-year institution teaching LLM internals — viemccoy · 2026-11-11
- The same people cite an existential AI race with China and want looser GPU export controls — austinc3301 · 2026-11-11