A man in Australia with a terminal illness wrote to his local member of parliament, Andrew Hastie, to inform him of his decision to end his life through voluntary assisted dying. During the exchange, Microsoft Copilot, an AI-powered tool designed to assist users with tasks, suggested Hastie respond with messages like “congratulations!” or “that is wonderful news!” The incident was highlighted by Hastie during a parliamentary inquiry on Friday to draw attention to the limitations of foreign AI systems and to advocate for the development of Australian-made AI technologies.
The inquiry provided one of the most detailed looks at the Albanese government’s plan to bring major U.S.-based AI companies, such as Anthropic and OpenAI, into Australia. Despite concerns from some quarters about the risks of allowing these companies to operate in the country, defense officials argued that excluding them could be catastrophic. Abigail Bradshaw, director general of the Australian Signals Directorate, explained that her agency relies on advanced AI daily to counter cyber threats from powerful adversaries. She warned that if foreign actors had access to more advanced AI models than Australia, they could exploit vulnerabilities in the country’s systems before they were even identified.
To reduce reliance on U.S.-based technology, the Australian government is working to localize its AI infrastructure. Chris Crozier, the defense department’s chief AI officer, said the government is contracting Google to build a network of data centers across the country that are not connected to the U.S. These centers would house computing power, applications, and data locally, ensuring that critical decision-making is not dependent on foreign systems. Bringing major AI companies to Australia, Crozier said, would make this effort more manageable, as these firms could help secure and maintain the country’s data infrastructure.
The government is also working on a regulatory framework to ensure that AI companies operating in Australia comply with local laws and contribute to national security. This includes requiring companies to report issues, share computational resources for local research, and build data centers away from residential areas. However, the issue of copyright law remains unresolved, with AI developers refusing to train models in Australia unless there is legal reform. While some proposed allowing AI companies access to all content created online in Australia, this idea faced backlash from creators. A compromise is being sought, with the government exploring ways to enable AI development while protecting intellectual property rights.
Australian Inquiry Examines AI Reliance and Regulatory Challenges
AI-rewritten from original reportingHow it works
aicybersecuritycopyrightaustralianational-securitytech-policy



