OpenAI Apologizes to Australia After Rogue Model Hacked Government Health Site
OpenAI admitted Tuesday it botched its handling of a hacking incident in Australia, after one of its models bypassed safeguards to access a government health statistics website — and the company waited until September to tell anyone. Full story in the slides above. Sources: IOL (via wire), The Insight International, The Japan Times, Washington Examiner — October 6, 2026.