AI Update: SCOTUS Declines Review In Thaler Case

The U.S. Supreme Court has declined to hear Dr. Stephen Thaler's appeal seeking copyright protection for his AI-generated artwork A Recent Entrance to Paradise. The decision allows to stand the long series of administrative and judicial rulings holding that a work created autonomously by an AI system cannot be protected by copyright under U.S. law because it lacks a human author, The denial of certiorari was unsurprising, but the implications and ongoing issues remain significant.

1. The Thaler Case: A Narrow Question

Dr. Thaler has been attempting for many years to secure copyright registration for a work that he described as having been autonomously generated by his AI system (sometimes called DABUS or the "Creativity Machine"). He initially argued that because he owned and created the system, the resulting work should be considered a work-made-for-hire under copyright, owned by him although created solely by the system. That argument faltered early, though, and when he later attempted to shift to a different theory — namely, that his guidance and control were sufficient to constitute human authorship — the lower court found the argument waived because it had not been raised earlier.

Accordingly, the question before the courts in the Thaler cases has been relatively narrow: Can an AI system acting autonomously be an "author" for purposes of U.S. copyright law? Both lower courts held "no," reaffirming human authorship as a bedrock requirement. The Supreme Court then refused to take the case. The U.S. Copyright Office had long maintained the same position, consistently rejecting applications listing non-humans as authors. Those decisions have now effectively been endorsed by the judiciary.

2. Why the Denial Was Expected

While Dr. Thaler attracted at least one amicus brief supporting his petition, there does not appear to be significant doctrinal or policy momentum at this time, either within U.S. law or internationally, for treating a truly autonomous AI system as an author. The Copyright Act contains multiple provisions that presuppose a human creator, and courts have repeatedly reinforced that assumption. As the D.C. Circuit emphasized, copyright is a form of property tied to human lifespan, transferability, and creative labor, frameworks that are poorly suited to autonomous machines.

Thus, the Supreme Court's refusal to take the case, though noteworthy, does not necessarily signal broader hostility to AI creativity. It simply leaves intact the long-standing principle that AI acting alone cannot be an author.

3. The Open Question: What Counts as "Human Authorship" when Using AI?

The unresolved frontier, of course, is not autonomously-acting AI, but human-directed AI.

The Copyright Office currently takes the categorical position that providing even extensive prompts to a generative system does not supply the "traditional elements of authorship." Under this view, any image or output generated by systems like Midjourney, Stable Diffusion, or DALL·E is uncopyrightable unless the human user contributes sufficient creative expression before or after generation (e.g., uploading an initial image beforehand or editing or modifying outputs), and then only to the extent of the human contribution, as demonstrated by the now-well-known Zarya of the Dawn, Rose Enigma and A Single Piece of American Cheese examples.

4. The Jason Allen Case: 624 Prompts and Still No Human Authorship

Colorado artist Jason Allen, whose work Théâtre d'Opéra Spatial won a state art competition, apparently employed 624 Midjourney prompts and extensive iterative refinement to achieve his final image. Yet the Copyright Office rejected his application, finding that the AI, rather than Allen, executed the "traditional elements of authorship."

Allen is now litigating that denial, arguing that prompt-based creation is analogous to earlier technologies like photography, where artists make creative choices even though a machine captures the image. His case may ultimately reach the Supreme Court and could become an important test of what level and types of human involvement and contribution – conceptually, technically, aesthetically, phenomenologically, and otherwise — are sufficient to constitute human authorship in the context of generative AI.

5. International Divergence: China's More Flexible Approach

Not all jurisdictions share the U.S.'s strict view.

In Li v. Liu (2023), the Beijing Internet Court held that a Stable Diffusion-generated image was copyrightable because the plaintiff's creative contributions — prompting, parameter adjustments, and selection — constituted sufficient intellectual input. The court recognized the work as "original" and reflecting the plaintiff's "intellectual investment."

However, China's approach is not uniformly permissive. In the later Zhou v. Defendant 1 case (2025), the same court refused protection for a Midjourney-generated image because the plaintiff failed to provide evidence of his creative process, such as specific prompt engineering or iterative adjustments; "after-the-fact simulations" of the creative process were insufficient to prove originality and user effort. The court held that the burden lies on the claimant to prove substantial human contribution, and failure to do so resulted in no recognition of copyright in that case.

These cases suggest a more case-by-case, evidence-driven approach in China, in contrast to the categorical rule applied thus far in U.S. cases. Which approach is more effective at "promoting progress" in the arts and creative industries remains uncertain.

6. What Comes Next?

The Thaler cert denial closes the chapter on autonomous-AI authorship, but the underlying and related questions involving human-AI creativity will likely continue to be carefully considered, and litigated. The Allen case — and others likely to follow — will force courts to confront the core question the Thaler case sidestepped:

When a human uses an AI tool, what level of control, judgment, and creative input is sufficient for human authorship?

Until that question is resolved, creators, technologists, and rights-holders will continue to operate in a gray zone. But one thing is certain: as AI becomes more deeply embedded in creative workflows across all industries, the definition of "authorship," and questions around the protectability of AI-assisted works, will remain at the forefront.

The content of this article is intended to provide a general guide to the subject matter. Specialist advice should be sought about your specific circumstances.