During a parliamentary inquiry on Friday, Australian MP Andrew Hastie recounted an episode where Microsoft Copilot generated inappropriate response suggestions when he received correspondence from a constituent with a terminal illness planning to pursue voluntary assisted dying. The AI system suggested replies including “congratulations!” and “great to hear from you,” demonstrating significant limitations in content filtering and contextual understanding.
Hastie used the incident to support calls for domestically controlled artificial intelligence infrastructure. The inquiry examined the Albanese government’s strategy to attract major AI developers Anthropic and OpenAI to Australia despite ongoing concerns about accelerating AI development. Defence and security officials made the case for such partnerships, arguing that excluding advanced AI capabilities could pose national security risks.
Abigail Bradshaw, director general of the Australian Signals Directorate, warned that adversaries with access to more advanced AI models could identify and exploit vulnerabilities before Australian cyber agencies could detect them. Defence department chief AI officer Chris Crozier noted that the government is already working to localize computing infrastructure through contracts with Google to build interconnected datacentres across the country. Officials argued that hosting major AI companies on Australian soil would provide government oversight, earlier access to advanced models, and greater technical integration with frontier developers.
The government is developing regulatory standards for companies establishing operations in Australia, including requirements for reporting breaches, reserving computational power for local research, and community engagement around large datacentre projects. A significant outstanding issue involves copyright law reform, which AI developers say is necessary for training models locally. The proposal to grant default access to Australian-created online content has drawn opposition from creative industries, who argue companies should negotiate licensing agreements instead.
Government officials acknowledged the urgency of decisions around AI development timelines. Bradshaw cautioned that delays in establishing favorable conditions could prompt companies to locate elsewhere, potentially limiting Australia’s influence and access to cutting-edge AI capabilities.
Article Attribution | Read More at Article Source
Article summary produced by Claude AI