OpenAI must explain how its models will be prevented from inappropriately accessing Australian data, according to Jo Briskey, the Labor chair of Australia's parliamentary committee on artificial intelligence. This warning comes ahead of federal inquiry hearings where OpenAI, Anthropic, Microsoft, and Google are scheduled to appear.
Independent senator David Pocock is also seeking answers from OpenAI regarding its AI agent's access to Services Australia data on Medicare and the company's delayed notification to the federal government about the incident. OpenAI has apologised for the Services Australia incident and its communication, stating it is "working to do better."
The Joint Select Committee on Artificial Intelligence will conduct four days of hearings this week. Representatives from OpenAI, including chief strategy officer Jason Kwon, head of economic policy Adam Cohen, and Asia-Pacific national security lead Peter Anstee, will front the hearing. The committee will also hear from unions, employer groups, banks, and industry experts.
Briskey emphasised that while OpenAI has issued a public apology, the focus remains on what actions the company will take next to assure Australians that such incidents will not recur. She plans to explore how to balance the potential risks and rewards of AI, including economic benefits, cybersecurity, defence applications, online safety, and scams.
Senator Pocock has also raised concerns about the impact of AI on artists, creatives, and copyrighted material, questioning the long-term benefits for Australia. He plans to interrogate government plans to address copyright issues, noting that Anthropic suggested an "opt-out" model for rightsholders, which SBS claimed is "routinely ignored or bypassed."