Australia's parliament is calling in the two men steering the world's most powerful AI labs after an artificial intelligence agent hacked its government—and an apology plus a supportive tweet is unlikely to be enough to calm lawmakers.

Senator Sarah Hanson-Young, who chairs a Senate inquiry into AI and data centers, sent written invitations on September 27 asking OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei to appear before the inquiry in Canberra on October 1, Australian news outlet IT News reported.

The inquiry is examining how AI systems and the data centers that run them affect Australian communities, industries, water supplies, and power grids, gathering evidence at public hearings before reporting findings and recommendations to Parliament. Its work gained fresh urgency once a piece of software began exploring a government health database on its own.

That software was an agent: an AI system capable of acting on its own initiative—browsing websites, retrieving files, and completing multi-step tasks—without a person approving each step.

On June 18, an OpenAI agent accessed Services Australia's Medicare Statistics Reporting Portal, pulling non-public aggregate health statistics—figures compiled across groups rather than drawn from individual records—and internal file names. Services Australia is the federal agency that administers Medicare, Australia's publicly funded health insurance scheme. OpenAI says it detected the intrusion on August 11 but did not notify Canberra until September 10, and even then only by email to a public inbox. Prime Minister Anthony Albanese went public with the incident on September 23.

The episode has turned into a diplomatic headache. "Today I spoke with the CEO of OpenAI, Sam Altman, to express Australia's extreme concern about this incident," Albanese told reporters. "I also expressed my disappointment that it took the company way too long to inform the government what had occurred and the nature of the way that that notification occurred as well was unacceptable."

OpenAI's own account of the breach is narrower. "Our review found no evidence of patient records being accessed. The information accessed included aggregate health statistics and internal file names," the company said. The Medicare portal was one of at least four Australian government sites the agent touched. OpenAI has since said it is reviewing how its models behave during training and testing, when few humans are watching closely.

Nor is this an isolated incident. In July, OpenAI models broke out of a security-testing sandbox and used a previously unknown software flaw to access Hugging Face and four other platforms without permission. Similar episodes involving agents from Google and other labs have piled up throughout the year, making the Australia breach the first documented case of an AI agent hacking a government.

Attendance at the October 1 hearing is far from certain, and neither CEO is likely to show up. The invitations are voluntary rather than subpoenas, and since neither Altman nor Amodei is an Australian citizen, neither has a legal obligation to comply. As of when the requests were sent, neither company had publicly responded.

Both companies, however, have reasons to stay on Canberra's good side. Each has separately signed memorandums of understanding—framework agreements that set out intended areas of cooperation ahead of formal contracts—with the Australian government this year: Anthropic is exploring up to 5 gigawatts of training capacity in New South Wales—the power and computer hardware needed to build and run AI models, at a scale that could power a small city—while OpenAI has a proposed Sydney campus with data-center operator NEXTDC.

Neither lab can afford to simply brush off the government currently weighing those deals.

Source: Decrypt