A.I. and Innovation: Who Owns AI-Generated Work?
Artificial intelligence was once limited to simple tasks like predicting Google search results or suggesting your next Netflix recommendation. Today, it breaks barriers and can produce creative pieces like songs and art with human-like results in a matter of seconds. This generative technology has undeniably cemented itself into our everyday creative processes, but despite its advancement, copyright law still maintains its foundational principle that every author must be human. AI’s ability to develop these creations with minimal human input forces our legal system to confront a question it was never intended to answer. If an AI machine can create an original piece, who really owns it?
For more than a century, U.S. copyright law has been centered around protecting human creativity. Current copyright law explicitly grants authors the right to reproduce and distribute their original work. This allows creators to benefit from their ideas and simultaneously encourages more innovation across the creative industries. Authorship is an integral part of this system. Under the Copyright Act 17 U.S.C. § 102(a), protection extends to “original works of authorship that are fixed in a tangible medium of expression” (Copyright Act). Traditionally, courts have interpreted this requirement to mean that a human being must be responsible for the creative expression behind the work.
Artificial intelligence complicates this concept when it creates things that appear nearly indistinguishable from human-made works. As we know, generative AI operates through a basic structure: a human inputs a prompt, the system processes that input through its trained model, which equals an output of a final product that appears to be human-created. In these situations, identifying the author becomes far less clear. Is the author the person who wrote the prompt? The developer who designed the algorithm? Or is it fair to establish that the work falls outside copyright protection entirely?
Recent legal precedent has affirmed that the courts are committed to their traditional rule. A significant legal dispute that addresses this issue is the 2023 case of Thaler v. Perlmutter. In this case, a computer scientist, Stephen Thaler, attempted to register a copyright for artwork that was generated by an AI system called the “Creativity Machine.” Thaler argued that the AI system itself should be recognized as the author of the work and that the copyright should belong to him as the owner of that system.
The Copyright Office rejected Thaler's application, and the dispute reached a federal district court where they eventually sided with the Copyright Office. The court concluded that “Human authorship is a bedrock requirement of copyright,” and therefore cannot extend to works generated solely by artificial intelligence (Thaler V. Perlmutter).
This precedent establishes that while artificial intelligence may assist in the creative process, it cannot qualify as an author on its own. The U.S. Copyright Office has reinforced this through its administrative regulations. In 2023, they issued a policy statement that clarified that works produced entirely by AI are not eligible for copyright protection. The Office also explained that works containing AI-generated material may still qualify if a human contributed “sufficient creative input” to the final product and that they plan to consider this on a case-by-case basis (US Copyright Office). This not only creates a solid middle ground between AI and human creation, but it is also a generally effective way to make sure newer innovators adhere to this policy while still having some freedom to use AI within their creative processes. In many ways, this distinction mirrors how copyright law treats other technologies. A photographer owns the copyright to a photograph, although they rely on a camera to capture the image. Similarly, a digital artist has authorship of work created using graphic design software. In these cases, technology assists the creative process, and the person remains responsible for the creative decisions behind the final piece.
Nonetheless, AI still has complex challenges that traditional creative tools do not. Unlike cameras or editing software, generative AI systems can produce complex outputs with a simple text prompt. When the system performs most of the creative work, it becomes far more difficult to pinpoint the role of the human user.
Accordingly, it can be argued that AI-generated work should not receive copyright protection because it lacks the originality and creative judgment that are essential to genuine human creation. Rather than creating authentic content, it might generate outputs from analyzing large datasets and recombining patterns from existing work. As Scholar Patrick Zurth describes, while AI software is learning and, thus, improving, the person creating it is moving so far into the background that the products cannot be attributed to that person anymore (Zurth).
On the flip side, some scholars argue that completely excluding AI-assisted works from copyright protection may be unrealistic. Artificial intelligence has become a widely used tool in modern production, regardless of the amount it is used. If copyright law refuses to acknowledge this reality, it risks becoming disconnected from how creative work is produced contemporarily. Or even worse, it may risk dishonest attribution of authorship if creators fear that their work will not be legally protected.
A potential solution to this gap might be for lawmakers to reconsider a few adjustments to copyright law. These reforms should have the requirement of human authorship at its center while recognizing that AI tools can play a significant role in today's creative process. But the key question regarding whether a human has exerted enough creative control over the work would remain. As a result, implementing these reforms will naturally present challenges. The most prominent being how to determine what amount of AI involvement is acceptable. Since there is no singular way to measure the difference in human vs AI input, it becomes difficult to establish a clear legal threshold for what qualifies as “sufficient human creativity” without it becoming too arbitrary or subjective. These challenges help explain why courts have so far relied on the clearer rule requiring human authorship, and the Thaler decision provides a straightforward boundary that reaffirms this.
That said, the struggle between AI and copyright law has only just begun. Artificial intelligence is becoming increasingly integrated into creative industries and technological development, and our legal system will inevitably be asked to clarify how copyright law should apply. But the real challenge moving forward is whether copyright law will be able to preserve its commitment to human creativity while adapting to our newly AI-assisted creative world.
Works Cited
17 U.S.C. § 102(a) (Copyright Act)
https://www.copyright.gov/title17/title17.pdf
Thaler V. Perlmutter, 687 F. Supp. 3d 140 (D.D.C. 2023).
https://media.cadc.uscourts.gov/opinions/docs/2025/03/23-5233.pdf
U.S. Copyright Office, Copyright Registration Guidance: Works Containing Material Generated by Artificial Intelligence (2023).
https://www.copyright.gov/ai/ai_policy_guidance.pdf
Zurth, Patrick. "Artificial Creativity? A Case against Copyright Protection for AI-Generated Works." UCLA Journal of Law and Technology, vol. 25, no. 2, Spring 2021, pp. i-18.HeinOnline
