Supreme Court Silence on AI Authorship Leaves Human Requirement Intact

The Supreme Court's March 2 denial of certiorari in Thaler v. Perlmutter marks the end of Dr. Stephen Thaler's years-long campaign to secure copyright protection for purely AI-generated works. But the Court's silence does more than close one door: it leaves the entire framework for AI creativity in legal limbo.

What Was Actually Decided

Thaler sought copyright registration for "A Recent Entrance to Paradise," a visual artwork created entirely by his AI system DABUS without human intervention. He listed the AI as sole author, naming himself as owner by virtue of creating the machine. The Copyright Office refused registration, the D.C. District Court affirmed in 2023, and the D.C. Circuit upheld that ruling in 2025.

The appellate court's reasoning was structural: the Copyright Act presupposes human authorship through provisions tying copyright duration to human lifespan, transfer mechanisms that require legal persons, and the constitutional language of "authors" in the Progress Clause. Judge Patricia Millett wrote that "adhering to the human-authorship requirement does not impede the protection of works made with artificial intelligence" - a crucial distinction that preserves copyright for AI-assisted works while denying it to autonomous machine output.

The Broader Stakes

Thaler's petition asked whether "works outputted by an AI system without a direct, traditional authorial contribution by a natural person could be copyrighted." That framing deliberately avoided the messier middle ground where most AI creativity actually happens. The case presented the cleanest possible test: zero human creative input versus traditional human authorship.

The federal government urged denial, noting that copyright duration - life plus 70 years - cannot sensibly apply to machines. They argued that extending copyright to non-human creators "would require legislative action," not judicial interpretation.

Conservative groups filed amicus briefs arguing that AI-generated works earning copyright protection could "chill speech among creators" facing infringement claims from works lacking traditional fair use protections premised on human consent. Academic amici pushed the other direction, warning that rigid human requirements might stifle innovation.

The Practical Impact

The denial leaves businesses with clearer guidance on purely autonomous AI output: it gets no copyright protection. Companies building AI systems that generate creative content without human direction cannot rely on copyright to protect that output from copying. This affects generative AI companies whose models produce standalone creative works and any business using AI for automated content creation.

The decision creates no roadmap for the vast middle territory of AI-assisted creativity. The Copyright Office's January 2025 guidance states that "prompts essentially function as instructions that convey unprotectable ideas" and current AI technologies don't offer enough control to qualify human prompters as authors. Yet the Office acknowledges courts must "provide further guidance on the human authorship requirement as it applies to specific uses of AI."

Historical Parallel

This mirrors the Supreme Court's approach to photography in Burrow-Giles Lithographic Co. v. Sarony (1884). The Court faced arguments that photographs couldn't be copyrighted because they were "the product of a machine." Justice Samuel Miller rejected that reasoning, finding copyright protection in the photographer's "intellectual conception" and artistic choices - the human creativity that directed the machine's output.

Today's AI presents the inverse challenge: when is the human input sufficient to cross the authorship threshold? The Copyright Office cited Sarony in denying Thaler's application, but that precedent offers limited guidance for determining when AI assistance becomes AI authorship.

What Comes Next

The Supreme Court's parallel denial in Thaler v. Vidal - where Thaler sought patent protection for AI inventions - establishes consistent treatment across IP law: only humans can be inventors or authors. This alignment suggests deliberate judicial restraint, leaving complex AI policy questions to Congress and agencies.

The USPTO continues expanding its AI engagement. Its Artificial Intelligence Search Automated Pilot Program (ASAP) was extended through June 2026 with waived fees and doubled capacity to 3,200 applications. Director Katherine Vidal has described ASAP as "the first of many planned AI pilots in patent examination."

The more immediate pressure comes from ongoing copyright litigation where AI companies face claims over training data scraping. Cases like Andersen v. Stability AI and NYT v. Microsoft present fact patterns far more complex than Thaler's binary test - human plaintiffs alleging their copyrighted works were used without permission to train AI systems that compete with human creators.

Thaler's failed petition asked the wrong question at the wrong time. The real battles over AI and copyright aren't about machine authorship; they're about human authors' rights when their works become training data. Those cases are percolating through district courts now, and unlike Thaler's clean hypothetical, they involve billions in commercial stakes and established copyright owners with standing to sue.

The Supreme Court's silence on AI authorship may prove temporary. Its message is clear: if Congress wants machines to be authors, Congress must say so explicitly.