From Tokyo: There is a particular irony in the fact that one of the most consequential intellectual property rulings of the generative AI era was delivered not with a thunderclap of judicial argument, but in near-silence. On Monday, the United States Supreme Court simply declined to hear the case, letting stand a string of lower-court decisions that say, plainly: machines cannot be authors.
The case belongs to Stephen Thaler, a computer scientist from St. Charles, Missouri, who applied for a copyright in 2018 over a piece of visual art his AI system, DABUS, generated autonomously. The image, titled "A Recent Entrance to Paradise," depicts train tracks entering a portal, surrounded by what appears to be green and purple plant imagery. Thaler listed DABUS as the creator. The US Copyright Office rejected his application in 2022, finding that creative works must have human authors to be eligible for copyright.
A federal judge in Washington upheld the office's decision in 2023, writing that human authorship is a "bedrock requirement of copyright." On 18 March 2025, a three-judge panel of the US Court of Appeals for the District of Columbia then determined the Copyright Office had correctly denied Thaler's claim. The Supreme Court's Monday refusal to intervene leaves that appellate ruling as the final word, for now.
The case is not simply an American curiosity. President Donald Trump's administration had urged the Supreme Court not to hear Thaler's appeal, arguing that while the Copyright Act does not define the term 'author,' multiple provisions of the act make clear that the term refers to a human rather than a machine. That bipartisan consensus across administrations and courts gives the ruling considerable weight as a global signal.
Thaler's legal team argued the stakes were enormous. His lawyers warned that even if the court overturns the Copyright Office's position in a future case, it will be too late, and that the office will have irreversibly and negatively impacted AI development during critically important years. That argument deserves serious engagement. The generative AI sector is moving at a pace that copyright law, designed for an era of human pencils and paintbrushes, was not built to handle. Companies investing heavily in AI-assisted creative production have a legitimate interest in knowing whether their output can be legally protected.
The DC Circuit's reasoning also carries its own internal logic. The court pointed out that machines, including the Creativity Machine, do not respond to economic incentives, so the human authorship requirement does not result in less original work, as the incentives for human creators remain intact. Copyright law, in this reading, was always about rewarding human endeavour, not production volume. That is a defensible position, and one broadly aligned with how Australia's own framework operates.
Under the Australian Copyright Act 1968 (Cth), copyright protection is granted to original works that originate from the 'author' and are expressed in a material form. The term 'author' refers to a human author, as AI was not a consideration when the Act was written. Australia is therefore facing the same foundational question as the United States, but without a definitive court ruling to anchor policy.
The government established the Copyright and Artificial Intelligence Reference Group (CAIRG) to consider whether and when AI-generated works should receive copyright protection. The government has asked the group to explore encouraging fair, legal avenues for using copyright material in AI through licensing arrangements, improving certainty on the application of copyright law to AI-generated material, and exploring avenues for less costly enforcement in relation to AI-related infringement. The Attorney-General has so far not provided a specific timeline for when AI copyright law reforms will be tabled to Parliament.
What the US ruling does not resolve is perhaps the harder question: what happens when a human and an AI collaborate? A wide swath of middle ground remains for works that involve some level of human and machine cooperation in the creative process, and harder questions remain as to what level of human influence might be necessary to qualify for copyright protection. A graphic designer who spent forty hours refining AI-generated imagery occupies very different legal territory from someone who typed a single sentence into a chatbot. Both are currently without clear guidance in most jurisdictions.
For the creative industries across the Asia-Pacific, this is not an abstract debate. Musicians, illustrators, game developers and filmmakers are already using generative tools as part of their everyday workflows. The absence of copyright protection for AI-assisted work does not just affect those creators' commercial interests; it creates incentives for corporations to rely on unprotected AI output rather than commission human artists, which is precisely the outcome the law's defenders say they are trying to prevent.
The Supreme Court's silence, then, is not a resolution. It is a signal to legislators that the courts will not do this work for them. In Australia, that means the CAIRG process carries more weight than ever. Getting the balance right, protecting human creative labour while giving AI-assisted creators reasonable certainty, is a genuine policy challenge without an easy answer. Reasonable people disagree about where the line should fall. The task now is to draw it deliberately, before the industry draws it for us.