US Tech Groups Challenge Australia Over Proposed AI Rules

OpenAI and Anthropic have urged the Australian government to reconsider its position on the use of copyrighted Australian content for training artificial intelligence models. The companies made their arguments in submissions to a parliamentary inquiry examining Australia’s proposed framework for regulating AI. Australia is preparing broader AI and data-centre regulations, while maintaining that existing copyright protections should continue to apply to AI companies.

Anthropic, the company behind the Claude AI assistant, proposed a limited system under which AI companies could obtain conditional approval to train models using Australian creative works. The company acknowledged that broad copyright exemptions had not received sufficient support from the Australian government, but argued that a narrower arrangement could potentially protect creators while allowing AI development to continue.

OpenAI also called for what it described as a balanced copyright framework. The company said AI models should be able to learn from publicly available information while giving copyright holders opportunities to collaborate with AI developers. Both companies linked their positions to planned investments in Australian AI infrastructure, including data-centre projects involving OpenAI and Anthropic.

The dispute reflects a wider international debate over whether AI companies should be allowed to use books, photographs, music, videos, social-media posts and other creative works to train commercial AI systems. Australian creators and rights organisations have raised concerns about their work being used without consent or compensation. At the same time, the government has been considering reforms that could provide AI companies with greater legal certainty around access to online material.

The issue is expected to remain part of Australia’s broader AI policy debate as lawmakers consider how to balance technological development, investment, copyright protection and the interests of creators. The parliamentary Joint Select Committee on Artificial Intelligence is scheduled to report its findings in November. Until then, the competing proposals from AI companies, creators and policymakers are likely to remain under consideration.