The Algorithm's Appetite: Australia at the Epicenter of AI's Copyright Conundrum

The global music industry is sounding the alarm as Australia finds itself uniquely positioned in a fierce battle over artificial intelligence and copyright law. According to the International Confederation of Music Publishers (ICMP), a peak global body representing giants like Sony, Warner, and Universal, Australian artists including Kylie Minogue and Nick Cave have already had their works used to train AI models without compensation. This development unfolds as the Australian government faces accusations of being 'strong-armed' by tech companies lobbying to weaken copyright protections.

John Phelan, the ICMP’s managing director, starkly characterized these efforts as an attempt to “legitimise larceny.” He warned that of the “dozens of countries approached” by AI firms to water down their intellectual property laws, Australia appears to be the only nation seriously considering such a move. The tactic, Phelan suggests, is a direct economic threat: tech companies are telling policymakers, “if you don’t do this, we won’t invest.” This direct linkage of copyright reform to investment signals a potent, and perhaps coercive, leverage point being applied to national governments.

The gravity of the situation was underscored in July by Prime Minister Anthony Albanese, who publicly stated it would be “theft” if artists were not granted control over their work's use by AI and compensated for it. Yet, the discourse seems to have shifted. A parliamentary committee on artificial intelligence recently described the situation as a “stalemate,” acknowledging that AI models are already training on copyrighted material in the United States without compensating creators. The implied quid pro quo is clear: without changes to local copyright laws, these companies are reportedly withholding the training of “frontier AI models in Australia.”

This creates a precarious economic tightrope for Australia. On one side, the creative sectors — music, film, media, and news organizations — express significant concern over behind-closed-doors lobbying that could undermine their livelihoods. On the other, the lure of AI investment and technological advancement dangles, creating pressure to align with the tech industry’s demands. A joint letter from the inquiry’s Labor chair, Jo Briskey, and its deputy chair, Liberal senator Jonathon Duniam, recommended the government “examine current copyright settings” while still aiming to ensure creators retain control and are paid. This nuanced position highlights the deeply conflicted policy landscape.

Crucially, the ICMP has challenged the AI industry’s claim that its models avoid training on Australian content due to existing copyright rules. Evidence provided exclusively to Guardian Australia by the ICMP demonstrated that Google’s Gemini AI was capable of reproducing lyrics from Australian artists, including Kylie Minogue, G-Flip, and Nick Cave. This capability directly contradicts industry assertions, suggesting that Australian content is, in fact, already part of the AI training landscape, often via methods like compiling “spreadsheets containing millions of lines of music available on YouTube.” The implication is that tech companies are already benefiting from Australian creative works, even as they push for legal changes to formally sanction such practices, potentially without fair compensation.

The global context is paramount. ICMP’s tracking of AI companies’ efforts to alter copyright law across different jurisdictions paints a picture of a coordinated, worldwide push to establish a precedent favorable to AI development at the expense of creator rights. Australia's position as seemingly the only nation considering such significant legislative concessions could set a dangerous global standard, impacting how artists worldwide are compensated and how their intellectual property is valued in the age of artificial intelligence. The debate in Canberra is therefore not just local policy; it's a test case for the future of creative economies globally.