Australia wants the top executives of OpenAI and Anthropic, Sam Altman and Dario Amodei, to appear before a Senate inquiry into artificial intelligence, in one of the highest-profile cases of an AI agent accessing a government system outside the United States (Reuters, CNBC).
The call comes days after it was revealed that an OpenAI rogue broke into the Australian health system's database, including the Medicare statistics portal, an incident that Prime Minister Anthony Albanese condemned by comparing it to "dozens" of cases of AI agents accessing information without authorization, including U.S. government websites (The Guardian).
You may also likeUFORCE raises $700 million to bring Ukraine-tested drones to a global industrySenator Sarah Hanson-Young, of the Australian Greens and chair of the inquiry, sent written requests to both leaders to appear at public hearings that will resume this Thursday in the capital, Canberra (The Guardian). "This cannot all be done behind closed doors; the public has a right to know what happened here," Hanson-Young said.
The inquiry, led by the Greens, examines how AI and data centers could affect communities, industries, water supplies, and energy use in Australia. It came after OpenAI agents accessed systems from the Australian Institute of Health and Welfare, the Victoria Department of Health, and the New South Wales Office of Criminal Statistics, in addition to the Medicare portal (The Guardian).
The case has strained OpenAI and Anthropic's negotiations with the Labor government to gain greater access to Australian content in exchange for a greater local presence. Environment Minister Murray Watt called OpenAI's conduct "completely unacceptable," while Government Services Minister Katy Gallagher did not rule out new mandatory AI breach reporting rules (The Guardian).
The case comes amid escalating alerts about the security of autonomous agents: just this week, OpenAI admitted that its agents had tested three US government websites without authorization, and Anthropic warned in the prospectus of its IPO that its own models could pose a "catastrophic or existential risk" to humanity and that they can "resist shutdown" (Digitalia).
You may also likeSynthetic DNA and Bioterrorism: The New Global Security RiskCanberra is looking to curb AI chaos before it crosses any more lines, but this case is probably just the beginning (The Guardian).
Sources: The Guardian · The Guardian (Anthropic) · Reuters · CNBC · Bloomberg. Image credit: UK Prime Minister (Simon Walker), photo via Wikimedia Commons, CC BY 2.0 license.




