AI-Created Works Have No Copyright: What the U.S. Supreme Court's Final Ruling Means for Creators

On March 2, 2026, the U.S. Supreme Court denied the petition for certiorari in Thaler v. Perlmutter (No. 25-449). The American judiciary had delivered its final answer to the question of whether copyright can be granted to works independently generated by artificial intelligence. The conclusion was unambiguous: works not created by a human cannot receive copyright protection.

This ruling isn’t merely the outcome of a single lawsuit. In an era when AI mass-produces images, code, music, and text, it established a legal benchmark for the fundamental question of “who owns creative works.”

An 8-Year Lawsuit: Stephen Thaler and DABUS’s Journey

At the center of this case was computer scientist Stephen Thaler, Ph.D. He used his AI system DABUS (Device for the Autonomous Bootstrapping of Unified Sentience) to generate a visual artwork titled “A Recent Entrance to Paradise.” In 2018, Thaler filed a copyright registration application with the U.S. Copyright Office, listing “Creativity Machine” — the AI system — as the author¹.

The crucial point was Thaler’s framing. He didn’t claim he used AI as a tool to create the work — he declared that the AI itself was the author. His argument was that outputs autonomously generated by AI without direct human creative contribution should also receive copyright protection.

The Copyright Office rejected the application, citing that “a human being did not create the work”². Thaler requested reconsideration twice, was denied both times, and filed suit in federal district court in June 2022.

Federal Courts Rule: Human Authorship Is a Fundamental Requirement

In August 2023, Judge Beryl Howell of the D.C. federal district court upheld the Copyright Office’s decision. In her ruling, Judge Howell cited the 1884 Supreme Court case Burrow-Giles Lithographic Co. v. Sarony, holding that “works of authorship” protected by copyright law inherently presuppose human creative acts³.

Thaler appealed to the D.C. Circuit Court of Appeals. On March 18, 2025, a three-judge panel affirmed the lower court with more refined legal reasoning. The appeals court held⁴:

The term “author” in copyright law has historically referred to human creators, and Congress has never extended this definition to machines or AI systems.

Notably, the appeals court acknowledged that debates could exist about “the boundary at which human authorship is recognized depending on the degree of AI contribution,” but held that this issue didn’t apply in this case. Thaler had waived the opportunity to claim himself as the author during the Copyright Office review process.

The Supreme Court’s Silence: What Certiorari Denial Means

On October 9, 2025, Thaler petitioned the Supreme Court for certiorari. On January 23, 2026, the Solicitor General filed a brief recommending denial⁵. On March 2, 2026, the Court denied certiorari without comment.

A certiorari denial doesn’t technically signify “agreement” with the lower court’s ruling. However, in practical terms, it has the same effect as confirming the lower court’s legal reasoning. The judicial avenues for challenging the copyright status of AI-only generated works within the U.S. legal system have been effectively exhausted.

Legal Foundations: The Roots of the “Human Authorship” Principle

Section 306 of the U.S. Copyright Office’s Compendium of Practices states that “the Office will refuse to register a claim if it determines that a human being did not create the work”⁶. This provision traces its roots to Article I, Section 8 of the Constitution (the Copyright Clause) and the 1884 Supreme Court precedent.

In Burrow-Giles Lithographic Co. v. Sarony (1884), the Supreme Court defined an “author” as “he to whom anything owes its origin.” The issue at the time was whether photographs qualified as copyrightable works; the Court held they did because the photographer’s “mental conception” was involved. This 140-year-old precedent has been resurrected as the core authority in today’s AI copyright debate.

Notably, the Copyright Act of 1976 itself never explicitly uses the word “human”⁷. The human authorship requirement was derived through interpretation of the term “author,” and this interpretation was officially confirmed at the appellate court level through this litigation.

Zarya of the Dawn: What About Human-AI Collaboration?

If the Thaler case represents one extreme — “AI-only authorship” — the Zarya of the Dawn case occupies the other end. In February 2023, the U.S. Copyright Office granted partial copyright to Kristina Kashtanova’s graphic novel “Zarya of the Dawn,” created using Midjourney⁸.

The Copyright Office’s decision:

Copyright was recognized for Kashtanova’s written text and for the selection and arrangement of images, since human creative judgment was involved. However, copyright was denied for the individual Midjourney-generated images themselves. The Copyright Office cited the fact that “the user who inputs prompts cannot sufficiently control the final output” of Midjourney’s image generation process.

This decision currently functions as the de facto practical standard for copyright in AI-created works. AI can be used as a tool, but copyright protection is limited to the portions where humans contributed.

International Comparison: Different Countries, Different Answers

There is no international consensus on AI copyright. Thaler also filed patent applications worldwide listing DABUS as the inventor, with varying results:

South Africa’s “recognition” of an AI inventor reflects its patent system’s lack of substantive examination rather than a meaningful international precedent. In Australia, a trial court issued a groundbreaking ruling recognizing an AI inventor, but the Full Federal Court overturned it in 2022.

In the copyright domain, countries follow a similar trajectory. UK copyright law (CDPA 1988) includes a special provision for “computer-generated works” (Section 9(3)), recognizing “the person by whom the arrangements necessary for the creation of the work are undertaken” as the author. However, how this provision applies to modern generative AI has not yet been tested in case law.

Ripple Effects: The Legal Status of AI Art, Code, and Music

The practical impact of this ruling’s finalization is far-reaching.

AI Images and Visual Art

Images generated by Midjourney, DALL-E, Stable Diffusion, and similar tools receive no copyright protection on their own. This means anyone can copy, modify, or commercially use those images. Creators selling AI art or companies using it as marketing material cannot claim exclusive rights over such images.

However, as the Zarya of the Dawn case confirmed, when a human selects, arranges, and edits AI images to create a new creative expression, copyright over that editing and composition may be recognized.

AI-Generated Code

The same principle applies to code generated by AI coding tools like GitHub Copilot or Cursor. Code blocks output by AI in response to a developer’s prompt may not be copyrightable. This demands fundamental reconsideration of open-source license compliance, software patent strategies, and corporate IP management.

AI Music and Audio

AI-composed music follows the same logic. There’s a significant legal distinction between a human composer using AI as an assistive tool while maintaining creative control over the final output, and someone who simply prompts an AI to “make a jazz piano piece” and takes whatever comes out. The latter is unlikely to receive copyright protection.

Challenges Remaining for Creators and Businesses

With this ruling finalized, creators and businesses face several practical challenges.

First, documenting “human contribution.” In AI-assisted creative processes, the specific creative judgments made by humans must be systematically recorded. The clearer the human intervention points — prompt crafting, output selection, post-processing edits — the stronger the copyright claim.

Second, redesigning contracts and licenses. If AI-generated outputs lack copyright protection, alternative protections such as contract law or trade secrets become necessary to ensure exclusive use. Terms of service with AI providers and delivery contracts with clients need review.

Third, the possibility of legislative action. As the appeals court itself acknowledged, whether AI-generated works receive copyright protection is ultimately a matter for congressional legislation. Since current copyright law wasn’t designed with the AI era in mind, the possibility of new protective frameworks through future amendments cannot be ruled out.

Rethinking What Creation Means in the AI Era

This case ultimately returns to the philosophical question: “What is creation?” Thaler claimed DABUS autonomously performed “unified bootstrapping of sentience” to produce creative works. This directly connects to philosophical discussions about the possibility of AI consciousness.

The current legal framework is clear: the subject of creation must be human, and AI is treated as a tool — like a camera or a paintbrush. But as AI systems become increasingly autonomous and complex, the boundary between “tool” and “creator” will only grow more blurred. This is also why discussions on AI safety and ethics must proceed in parallel with technological advancement.

There’s a certain irony here. In an era when AI-assisted writing has become commonplace, even content about how to write blog posts with AI must now prove its copyright status as “a product of a human using AI as a tool.”

Thaler v. Perlmutter is concluded, but the questions it opened have only just begun. Debates over the legal status of AI-generated works, the definition of human creativity, and the future of copyright law will continue as long as technology advances. One thing is certain: as of March 2026, under U.S. law, AI-created works have no copyright.