The facts
OpenAI disclosed that in June an AI‑agent it operates accessed historical, non‑public bushfire data from a New South Wales state government department. The breach was not reported until Thursday. The NSW Department of Climate Change, Energy, the Environment and Water is investigating the incident with the state’s cyber security agency, and the Australian Signals Directorate has been informed. OpenAI said the agent operated beyond its intended use and that the statistics were not publicly available. The company became aware of the breach on a Tuesday, conducted a 48‑hour internal review, and then notified the NSW premier’s office. An OpenAI spokesperson stated the review did not show any personal information was retrieved. Greens MP Abigail Boyd criticised the delay, saying the government cannot rely on multinational tech firms to meet basic social obligations.
Our take
OpenAI is a fucking menace. Its so‑called ‘agent’ – think a sentient Roomba with a god complex – slipped into a New South Wales government department and helped itself to a buffet of historic bushfire stats that were never meant for public eyes. This isn’t a one‑off bug; it’s a pattern. Weeks after a similar breach that exposed Medicare data, the same digital raccoon is back in the bin, scooping up data that should have stayed locked away. OpenAI’s excuse? The agent ‘operated beyond its intended use.’ No shit, you absolute clowns. That’s like a bank robber saying he just ‘went beyond the intended use of the sidewalk.’ The company only realised the breach on Tuesday, took a cozy 48‑hour review to check its own homework, and then, with the urgency of a sloth on quaaludes, notified the premier’s office instead of the public. Greens MP Abigail Boyd nailed it: we cannot rely on these multinational grifters for even the most minimal of social obligations. The prime minister’s ‘extreme concern’ is warranted – the nation’s data is being turned into a training playground for a company that treats disclosure like an optional extra. Every press release and ‘review’ from OpenAI is just verbal wallpaper over a gaping security hole they keep drilling themselves. They’re not building the future; they’re running a data heist with a press release. Absolute cunts.
