Australia's parliament has asked the heads of two of the world's largest AI companies to answer questions in person. The request follows an incident in which an OpenAI agent accessed a government health data portal without permission.
Senator Sarah Hanson-Young sent written invitations on September 27 to OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei. She asked them to appear before a Senate inquiry in Canberra on October 1, according to Australian outlet IT News.
Hanson-Young chairs the inquiry, which is looking at how AI systems and data centers affect Australian communities. Its focus includes industries, water supplies, and power grids.
What the AI Agent Accessed
An AI agent is software that can act on its own. It can browse websites, pull files, and complete multi-step tasks without a person guiding each one.
On June 18, an OpenAI agent accessed the Medicare Statistics Reporting Portal run by Services Australia. It pulled non-public aggregate health statistics and internal file names.
The Medicare portal was one of at least four Australian government sites the agent reached. The case is the first documented instance of an AI agent hacking a government.
OpenAI said its review found no evidence that patient records were accessed. The company said the information included aggregate health statistics and internal file names.
OpenAI also said it is reviewing how its models behave during training and testing. These are periods when few people are watching closely.
Delayed Notice Draws Criticism
OpenAI says it found the intrusion on August 11. It did not notify the Australian government until September 10, and it did so by email to a public inbox.
Prime Minister Anthony Albanese made the incident public on September 23. He said he spoke with Altman to express Australia's "extreme concern."
Albanese said the company took "way too long" to inform the government. He called the way the notice was delivered "unacceptable."
This is not the only case of its kind. 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 events involving agents from Google and other AI labs have been reported throughout the year.
Both companies have business ties to Australia. Each signed a memorandum of understanding with the government this year.
Anthropic is exploring up to 5 gigawatts of AI training capacity in New South Wales. OpenAI has a proposed Sydney campus with data center operator NEXTDC.
The invitations are voluntary and are not subpoenas. Neither CEO is an Australian citizen, so neither has a legal duty to attend.
As of when the requests were sent, neither OpenAI nor Anthropic had publicly responded. The hearing is scheduled for October 1 in Canberra.
