- 0 Comments
- Content Marketing, Marketing
AI and Copyright in Australia: Why Jimmy Barnes, Paul Kelly and Other Musicians Are Speaking Out Artificial intelligence is changing the way we create, communicate and do business. From generating marketing content to assisting with research and design, AI is becoming an increasingly familiar part of everyday professional life. But behind its extraordinary capabilities sits a question that is becoming increasingly difficult to ignore. Who should benefit when artificial intelligence is trained using somebody else’s creative work? In Australia, that question has become the centre of a significant dispute involving some of the country’s most recognisable musicians, major technology companies and the federal government. Jimmy Barnes, Paul Kelly, Peter Garrett and Mark Seymour are among the artists calling on Prime Minister Anthony Albanese to protect Australian music from being used for AI training without appropriate permission or payment. As reported by The Australian on 9 November 2026, the musicians have voiced concerns about the use of Australian recordings in AI training datasets and the potential weakening of protections for creators. Their concerns extend far beyond the recording industry. The outcome of Australia’s copyright debate could influence how photographers, writers, designers, publishers, filmmakers and other creative professionals protect the value of their work. And for businesses increasingly adopting AI tools, it’s a conversation worth understanding. Why Are Australian Musicians Challenging AI Companies? On 9 November 2026, prominent Australian musicians renewed calls for the federal government to protect copyrighted music from unauthorised AI training. The artists argue that recordings of their work have appeared in datasets associated with AI training without their permission or compensation. One particularly striking detail involves Prime Minister Anthony Albanese’s own musical preferences. According to the musicians, recordings of all ten songs on the Prime Minister’s list of favourite Australian tracks have appeared in datasets used by AI developers. It’s an uncomfortable coincidence for a political leader who has publicly celebrated Australian music. The very artists responsible for creating songs that have become part of our national culture are now asking the government to ensure their work retains meaningful legal protection. Their concerns are part of a wider debate about how AI companies obtain creative material and whether the people who create it should be compensated. The issue also extends into advertising and production. In a separate investigation, the Australian Financial Review reported on 8 November 2026 that composers Antony Partos and Nigel Westlake had raised concerns about musicians being asked to recreate AI-generated tracks for commercial use. These developments highlight how AI copyright questions are increasingly affecting the everyday work of creative professionals. What Does AI Training Have to Do With Copyright? Generative AI models are developed using large collections of information and creative material. Depending on the system, those collections may contain books, photographs, articles, audio recordings, illustrations, computer code and other content. During training, AI models learn patterns from the material they process. Those patterns help them generate new responses or outputs. As explained in APRA AMCOS’s guide to AI and music, generative AI can analyse existing music to learn patterns involving melody, structure, rhythm and other creative elements. The legal question is whether using copyright-protected material in this process requires authorisation from the copyright owner, and under what circumstances. This differs from the question of whether an individual AI-generated output infringes copyright. Training and output are connected, but they raise distinct legal and commercial issues. For Australian creators, the concern is that their work may contribute to the development of valuable commercial products without a meaningful opportunity to consent, negotiate terms or receive compensation. For AI developers, the challenge is obtaining workable access to large volumes of material while navigating complex copyright arrangements. Both concerns are relevant to the policy debate. The disagreement centres on how to balance them. The Difference Between Opt-In and Opt-Out AI Licensing One of the most consequential questions is who should be responsible for obtaining or withholding permission. Under an opt-in approach, AI companies generally need to secure permission or an appropriate licence before using protected material. Under an opt-out approach, certain material may be available for use unless the copyright owner takes action to exclude it, subject to the rules of the particular proposal. That difference may appear technical, but it has significant practical implications. Consider an independent Australian musician who has spent years developing a catalogue of original recordings. Under an opt-in system, the musician or relevant rights holder has an opportunity to determine whether their work should be licensed. Under an opt-out arrangement, the musician may instead need to understand which systems are collecting their work, identify how to register an objection and monitor whether that objection is respected. For a large organisation with legal resources, those responsibilities may be manageable. For an independent artist, they could become a substantial additional burden. The same concerns extend to photographers, illustrators, writers and smaller creative businesses. The Australian Independent Record Labels Association (AIR) outlined its concerns about copyright reform proposals, arguing that the rights of independent music businesses and creators should be protected as AI policy develops. What Is the Australian Government’s Position? Importantly, Australia has not introduced a broad AI training exemption into copyright law. In an official announcement on 26 October 2025, the Albanese Government confirmed it had ruled out a broad text-and-data-mining exception that would allow AI developers to use copyrighted material without permission. The government has continued examining copyright policy in response to emerging AI technologies, including questions about licensing, transparency and enforcement. The Attorney-General’s Department’s Copyright and Artificial Intelligence Reference Group provides information about the consultation process and the issues under consideration. These policy discussions should not be confused with changes already enacted into law. Different proposals may involve different arrangements, and their practical consequences depend on the final legal framework. As at 9 November 2026, the broader debate remains ongoing. Can AI Innovation and Creative Rights Coexist? The debate is sometimes presented as a choice between embracing technological innovation and protecting creators. I don’t believe that framing is especially helpful. AI
Learn More