Australia is simultaneously courting AI investment and confronting AI failure, a combination that is forcing the Albanese government to draft rules across copyright, child safety, and judicial oversight that no peer democracy has yet attempted.
Why it matters
Prime Minister Anthony Albanese's visit to Apple headquarters in Cupertino this week produced two concrete policy signals: openness to an opt-out copyright model that would let creators withdraw their works from AI training rather than requiring companies to seek permission upfront3,5, and a push for global AI guardrails that he plans to advance at next week's United Nations General Assembly4,6. Those moves landed the same week Tasmania's Justice Department confirmed it would review past parole board decisions after AI-generated "fictitious" case law was used to justify a condition imposed on convicted murderer Susan Neill-Fraser1,7. ANALYSIS The juxtaposition frames Australia's emerging posture: attract the labs, protect creators, and clean up the messes that arrive when AI tools are deployed without safeguards.
The big picture
Albanese told reporters in Cupertino that "Australia can't do it alone" on internet safety and AI regulation2. He urged the United States and China, the two leading nations in the AI race, to strike a deal "in the interests of humanity" and ensure "humans remain in control". The call came as President Donald Trump moved in the opposite direction, stepping up efforts to fight regulation.
On copyright, the prime minister said writers, musicians, and other authors would retain control over their intellectual property and receive compensation for its use. He allowed for a model under which creators could opt out of having their works used for AI training, a shift from the current principle requiring companies to obtain permission before use. Anthropic and OpenAI have described Australia's existing copyright rules as one of the obstacles to investment in the country. ANALYSIS By floating an opt-out framework rather than the stricter opt-in default, Albanese is offering the labs a concession designed to unlock investment while preserving a nominal creator veto.
During a closed 45-minute meeting with Apple CEO Tim Cook, Albanese discussed the company's new parental control tools, which let parents control apps, content, and contacts their children interact with, impose screen-time limits, and detect images containing nudity, gore, or violence. Those tools will be made available in Australia. Albanese said Australia is developing national standards for AI and data centres covering safety requirements, environmental rules, and industrial rules, and expressed hope that the relevant legislation would be adopted early next year.
ANALYSIS The Tasmanian parole case supplies the domestic urgency that makes Albanese's international pitch more than aspirational. Tasmania's Justice Department confirmed it will review the extent to which AI may have been used to inform past parole board decisions. The review follows revelations that the parole board relied on a document in Neill-Fraser's case that cited case law described as "fictitious" and likely "AI hallucinations". A parole condition preventing Neill-Fraser from speaking to the media about her alleged innocence was removed earlier this week. Barrister Greg Barns SC said anyone involved in generating the fake case law should resign or be sacked.
The incident is a concrete example of the risk Albanese referenced in Cupertino when he said, "I certainly accept that there are risks and that's why we need to mitigate them". A hallucinated citation shaped a real legal outcome for a person convicted of murder, the kind of failure that converts abstract safety debates into political imperatives.
Australia's regulatory agenda now spans three distinct fronts: trade-offs with AI labs on copyright to draw investment, bilateral engagement with platform companies on child safety, and retrospective audits of AI use inside government. No single peer country is moving on all three simultaneously, though each element has parallels elsewhere.
What's next
Albanese will speak with like-minded leaders at the United Nations General Assembly about global rules to regulate AI development. Draft online-safety laws will be introduced to parliament later this year after consultation with social media companies, industry bodies, and civil society groups. The national standards for AI and data centres are targeted for legislation early next year. In Tasmania, Attorney-General Guy Barnett said the department would establish an independent review into the parole board.