The Australian Productivity Commission (PC) is examining whether technology companies should be exempt from existing copyright rules that prevent them from mining text and data to train artificial intelligence (AI) models. This inquiry, part of its interim report on “harnessing data and the digital economy,” uses copyright as a case study. It explores how Australia’s current regulatory framework can change to manage the evolving risks and opportunities presented by AI.
Scott Farquhar, co-founder of Atlassian, argues that the current rules are outdated in comparison to other countries. He believes that creating exemptions for text and data mining to train large language models could attract billions of dollars in foreign investment. However, this view has faced opposition from the Copyright Agency, a not-for-profit organization representing thousands of creators. The agency has proposed an alternative solution. It suggests that the government set up a new compensation scheme to pay creators whose work is used by tech firms to train their AI models.
Stephen King, one of the two commissioners leading the PC’s inquiry, highlighted the main question: “Are our laws fit for purpose with AI?” He explained that while it is a concern that AI companies might use copyrighted materials without proper compensation, the nation also has a strong interest in encouraging the development of AI tools that use this material. King compared the situation to the music industry, where collecting societies manage royalties for songwriters and performers. He suggested a similar model, authorized under competition laws, could be a possible solution for AI. The commission is also seeking feedback on other options, like expanding the “fair dealing” exemption in copyright law to include text and data mining, as long as AI companies have legally obtained and paid for the source material. A final recommendation is expected by the end of the year.
The commission’s interim report suggests that artificial intelligence could be the key to improving Australia’s productivity. PC modeling indicates that even under the most conservative estimates, AI could add $116 billion to the economy over the next decade. King stated that such an outcome could mean a $4,300 increase in the average Australian’s real wage within 10 years, with actual benefits possibly being much greater.
Despite the potential for significant economic benefits, the commission has recommended against creating an overarching, AI-specific piece of legislation. It warns that overly burdensome or clumsy regulation could stifle the transformative potential of the technology. This approach is likely to be well accepted by the government, as Assistant Minister for Productivity Andrew Leigh has shown support for this perspective. However, King stressed that the PC is not pushing for a “light-touch” approach to regulation. He explained that AI will make it easier for bad actors to engage in harmful conduct, though much of this conduct is already illegal. The commission’s recommendation is to identify where the harms exist and see if they are covered by existing laws. It also emphasizes that regulators need sufficient resources and powers to address these issues.
In addition to copyright and regulatory frameworks, the PC’s interim report also addresses privacy rules, which support a shift toward an “outcomes-based” approach, moving away from a “box-ticking” exercise where businesses might follow the letter of the law but not its spirit. This further highlights its focus on modernizing the regulatory landscape to fully take advantage of the opportunities presented by the digital economy.
