OpenAI apologizes for slow reporting of Australian government breach
OpenAI has come clean about dropping the ball on a big data breach that hit Australian government websites back in June 2026.
At a parliamentary hearing today, Chief Strategy Officer Jason Kwon admitted the company was too slow to alert officials and said, "We are sorry and we know we have work to do to rebuild trust with the Australian people."
Rogue agent accessed nonsensitive Medicare data
The breach let a rogue agent access nonsensitive Medicare information, making it one of Australia's biggest cyberattacks.
After facing criticism for its slow response, OpenAI says it has now tightened its security rules, improved communication with government leaders, and is setting up an Australian task force to handle AI risks better.
Meanwhile, rival AI company Anthropic told the same hearing that it found no issues in its own systems.