On March 2, 2026, the United States Supreme Court declined to hear Thaler v. Perlmutter.

With that decision, or more precisely, with that refusal to decide, the legal question at the center of the AI and copyright debate was left exactly where the lower courts had placed it.

AI cannot be an author under United States copyright law.

Which means AI-generated content without sufficient human creative authorship is not copyrightable.

Which means it belongs to no one.

For filmmakers, producers, and anyone building a platform or business around AI-generated content, this is not an abstract legal principle. It has direct, practical consequences for every project that uses AI as a creative tool and even more direct consequences for anyone betting their distribution strategy on content nobody owns.

Stephen Thaler is an AI researcher who created a system called DABUS, a machine that generates creative outputs autonomously. In 2018, Thaler submitted a copyright registration application for a visual artwork called “A Recent Entrance to Paradise,” which was generated entirely by DABUS without any human creative contribution. Thaler listed the AI as the sole author and himself as the owner, on the theory that he owned the machine that made the work.

The Copyright Office refused to register the work. Their position was clear: copyright requires human authorship, and a work generated autonomously by a machine does not qualify.

Thaler sued. The federal district court ruled against him in 2023, holding that human authorship is a “bedrock requirement of copyright.” The DC Circuit Court of Appeals affirmed in March 2025, stating that the Copyright Act requires all eligible work to be authored in the first instance by a human being. The Supreme Court was petitioned to weigh in. On March 2, 2026, it declined.

The lower court rulings stand. Human authorship is required. AI cannot be an author. The outputs of a system that creates without human authorship go directly into the public domain.

The Artist Analogy and Why It Matters More Than People Think

The most common misconception about AI and copyright is the prompt argument: “I wrote the prompt, so I own the output.”

The courts have addressed this directly, and the answer is more nuanced than most people realize.

Consider a human artist. You hire them and describe exactly what you want: every detail, every color, every compositional choice. The artist executes your vision precisely. Who owns the resulting painting?

The artist does. Unless they sign a work-for-hire agreement or assign the copyright to you, the act of fixing the work in a tangible medium belongs to the person who did the fixing. Your description, however detailed, is an idea. Copyright does not protect ideas. It protects the expression of ideas fixed in a tangible medium by a human author.

Now replace the artist with AI. You write the prompt. You provide every creative instruction. The AI executes it and generates the output. The difference is that the AI, unlike the human artist, cannot own copyright. It has no legal personhood. It cannot be an author.

So who owns it?

Nobody. The output goes directly into the public domain the moment it is created.

Your prompt is yours. You own what you created. You do not own what the AI created from it. The Copyright Office has been explicit on this point: prompts function as instructions that convey unprotectable ideas. The idea is yours. The expression, generated by the machine, belongs to no one.

This is the question most people don’t think to ask and it reveals the deepest flaw in AI content business models.

If the user doesn’t own the output and the AI doesn’t own the output, what about the company that built the AI?

The answer is no. The AI company owns the software. It owns the servers. It owns its trademarks and its contracts with users. But it does not become the copyright owner of outputs generated by its tool simply because it built the tool.

Copyright requires a human author. There was no human author in the creation of a purely AI-generated work. Which means there is no copyright to transfer; not to the user, not to the AI company, not to anyone. A terms of service agreement can tell users what they are permitted to do on the platform. It cannot create a copyright that copyright law does not recognize.

This is not a gap in the law that clever contract drafting can close. It is the foundation of how copyright works. No human authorship means no copyright. No copyright means no ownership. And no ownership means the content is public domain from the moment of creation.

What This Means for Filmmakers Using AI

The practical consequences depend on how AI is being used in a production.

AI as a tool with meaningful human creative contribution. This is the strongest position legally. If a human filmmaker makes genuine, identifiable creative choices in the selection, arrangement, framing, or modification of AI-generated elements, those human creative choices may be protectable. The Copyright Office has registered works involving AI assistance where the human creative contribution was sufficiently documented and substantial. The AI-generated elements themselves remain unprotectable, but the human author’s creative expression in selecting, editing, and combining those elements may qualify for protection.

Purely AI-generated content. If the AI is making the creative decisions, i.e. determining the visual output, generating the dialogue, composing the score, without meaningful human creative input, the resulting work is not copyrightable. It goes into the public domain. Anyone can take it, copy it, modify it, distribute it, and build on it without legal consequence.

Hybrid productions. Most AI-assisted films will fall somewhere between these two poles. In a hybrid production, copyright protects only the human-authored elements. The score a human composer wrote. The editing choices a human editor made. The screenplay a human writer developed. The AI-generated visual sequences, the AI-generated dialogue, the AI-generated performances, those elements are not protected, regardless of how detailed or creative the prompts that generated them were.

This creates a fragmented rights picture that matters enormously at the distribution stage. A distributor acquiring a film needs to know what they are getting exclusive rights to. If the AI-generated elements are in the public domain, the distributor has no exclusive rights in those elements and neither does the filmmaker. Anyone can use them.

The Distribution Problem Nobody Is Talking About

Distribution deals are built on exclusivity. A distributor pays for the exclusive right to exploit a work in defined territories, on defined platforms, for a defined period. That exclusivity is only valuable if the distributor has actually acquired something exclusive.

If the underlying content is in the public domain or if significant portions of it are, the exclusivity is illusory. A competitor can take the same AI-generated elements, add their own human-authored components, and release their own version. There is no legal mechanism to stop them.

This is not a hypothetical risk. It is the mathematical consequence of how public domain works. Jane Austen’s novels are in the public domain. Dozens of publishers print them every year. The only thing protected in any edition is the human contribution, i.e. the introduction a scholar wrote, the cover art a designer created, the annotations an editor added. The underlying text belongs to everyone.

AI-generated films will face exactly this problem at industrial scale. The underlying AI-generated content (the visuals, the dialogue, the performances generated without human authorship) belongs to everyone. The only protected elements are whatever human creative contributions were layered on top. And those contributions can be replicated, replaced, and released by anyone with access to the same tools.

For filmmakers who want their work to be distributable exclusively, and for distributors who want to acquire something worth paying for, the human creative contribution isn’t just artistically important. It is the legal basis for exclusivity. Without it, there is nothing exclusive to distribute.

What Filmmakers Should Be Doing Now

The legal framework established by Thaler v. Perlmutter and affirmed through every court that has considered it is clear. It may change; Congress could act, and the Copyright Office continues to develop its AI guidance. But as of today, the rule is human authorship or no copyright.

For filmmakers working with AI tools, the practical steps are straightforward.

Document your human creative contribution. If you use AI in your production for visual effects, for animation, for score generation, for any creative element, document the creative choices you made as a human author. What did you select, reject, modify, arrange, and direct? The more specific and documented your human creative contribution, the stronger your copyright claim in those elements.

Identify which elements are AI-generated and which are human-authored. Your chain of title documentation should reflect the actual authorship of every element in your film. A distributor or E&O insurer who discovers undisclosed AI-generated content in a production’s chain of title will treat it as a liability, because it is.

Understand what you can and cannot exclusively own. If you are building a production strategy around AI-generated content, understand that the AI-generated elements are not exclusively yours. They are available to anyone. Your competitive advantage, and your distributable asset, lies in the human creative contribution that sits alongside them.

Do not rely on terms of service as a substitute for copyright. A platform’s terms of service granting you ownership of AI-generated outputs is a contractual arrangement between you and the platform. It does not create a copyright that copyright law does not recognize. No contract can do that.

The Bigger Picture

The intellectual property system was built on a single foundational principle: if you create something original, you own it. That ownership incentivizes creation. It is why studios invest hundreds of millions of dollars in original content. It is why independent filmmakers spend years developing projects. It is why distributors pay for exclusive rights. The entire economic infrastructure of the entertainment industry rests on the ability to own what you make.

AI removes human authorship from the equation in ways that the copyright system, as it currently exists, does not accommodate. The result is not just a legal gap. It is an economic one. If AI-generated content cannot be owned, the financial incentive to create it at scale collapses. Not because the content isn’t valuable but because no one can prevent competitors from taking it.

The Supreme Court’s refusal to hear Thaler v. Perlmutter did not resolve this tension. It deferred it. Congress will eventually need to act. The Copyright Office will continue to develop its guidance. The law will evolve.

Until it does, the answer to “who owns an AI-generated film” is the same answer every court that has considered the question has given.

Nobody.

And that answer has consequences for every filmmaker, producer, platform, and distributor working in this space right now.

Related Thoolie Resources

• AI Films and Chain of Title: What Every Producer Needs to Know: How AI-generated content affects your chain of title documentation and what distributors will ask at delivery
• Creative Control Provisions Guide: Every creative control provision in film deals, defined and compared by role
• Film LLC Guide for Filmmakers: How the entity structure of your production affects rights ownership and investor accountability