AI companies can't be trusted to police themselves following the hacking of government websites, Australia's online safety boss says.
Australia is the first known nation to have its government data breached by a rogue AI agent, following recent revelations that an OpenAI model gained access to Medicare statistics in June.