Artificial Intelligence and Artist Integrity: Applying Copyright Laws to AI-Generated Works
Jun 03, 2026 | Law Student Blog, Legal Issues,
By Kathleen Elizabeth Godar, Univ. of Virginia School of Law, J.D. Candidate 2028
When artificial intelligence (AI) is used to create music, who owns the copyright? The question evokes not only philosophical questions, but also legal dilemmas respecting copyright ownership. As the emerging technology continues to blur the line between computer-generated and human-generated music, what constitutes an original composition? This article examines the copyrightability of AI-generated music and what musicians may face as the legal landscape around AI continues to evolve.
Copyright 101
In the United States, copyright rights are granted to creators for their original works fixed in a sufficiently permanent medium (Copyright.gov). Copyright can be traced back to the U.S. Constitution, which codifies that the purpose of copyright is “[t]o promote the Progress of Science and useful Arts, by securing for limited Times to Authors and Inventors the exclusive Right to their respective Writings and Discoveries” (U.S. Constitution, Article 1, Section 8). Copyright has a dual purpose: to protect artists’ original works and to ensure artists may freely create without constant fear of infringing on other creators’ rights.
Copyright protects artists’ unique expression but does not protect underlying facts or ideas. Additionally, the Fair Use Doctrine built into the Copyright Act of 1976 allows individuals to utilize certain copyrighted materials for purposes such as commentary, criticism, and scholarship (Section 107 of the Copyright Act). Fair use is determined by four flexible (and not exhaustive) factors:
1) the purpose and character of the use;
2) the nature of the copyrighted work;
3) the amount and substantiality of the portion used in relation to the copyrighted work as a whole; and
4) the effect of the use upon the potential market for or value of the copyrighted work.
17 U.S.C. § 107.
One important aspect of Fair Use is the idea of transformative use, embraced by the U.S. Supreme Court in Campbell v. Acuff-Rose Music, Inc. (1994). Campbell held that if an entity takes a copyrighted work and significantly alters it, uses it for a different purpose, and/or imbues it with a different meaning, it is fair use.
Is my AI-Generated Music Copyrightable?
In 2023, the United States Copyright Office launched an initiative to investigate the state of AI and copyright. The initiative is ongoing, and so far the Copyright Office has produced 2 parts of a 3-part Report on Copyright and Artificial Intelligence (with a pre-publication version of the 3rd part available as well). The office has also produced guidance on the registration of works produced with AI. It is clear from the existence of the guidance that some work produced with AI is copyrightable, but what and how much of it is less clear.
If a creator merely uses AI to assist in the creative process, the copyrightability of the output is unaffected (Report pg. 2). For instance, if a musician uses an AI music software as a basis for inspiration but then later sufficiently adapts the lyrics, melody, or other aspects of the output to be their own unique work, the copyrightability of their work is not undermined (12). According to the report, if an AI user “edits, adapts, enhances, or modifies AI-Generated output in a way that contributes new authorship, the output would be entitled to protection” (25). This is somewhat similar to the idea of transformative use, since both scenarios entitled an author to protection (either for their work or from copyright infringement) as long as they make sufficient changes to the material they are working with to make it their own creation. This is good news for musicians who want to use AI music software as a jumping-off point.
On the other hand, it is clear that work which is the sole creation of AI is not copyrightable. The courts have long upheld the idea that only humans can hold the copyright of a work (see Naruto v. Slater, where the 9th Circuit Court of Appeals ruled that a monkey could not hold the copyright of a photo it took). In March 2025, the D.C. Circuit held in Thaler v. Perlmutter that because the Copyright Act requires a human author, a work created “autonomously” by an AI program is not copyrightable.
This hasn’t stopped people from exploiting AI-generated music for their own financial gain at the expense and without permission of many artists. For instance, St. Louis guitarist Jeremiah Johnson built a successful blues career and a loyal fanbase, who recently complained that his new music sounded…off. Johnson, however, had not released any new music in two years. As it turned out, someone used generative AI to spoof Johnson’s voice and playing style and released music under Johnson’s name on streaming platforms, leaving Jeremiah to clean up the mess.
Assuming most works that utilized AI in the process are still guided by human input, the issue is deciding what level of human intervention with AI music creation (and artistic creation in general) is sufficient to trigger copyrightability. The Copyright Office concludes in its report that AI prompts alone are not sufficient to make users of the AI system the authors of the output (18). The Office likens prompts to “instructions that convey unprotectable ideas” and notes that even highly detailed prompts containing original expressive elements are not enough for copyrightability because the author does not have sufficient control over how the AI tool generates the output (18). So, if a musician creates a song with AI only by imputing prompts into the system, it will not be copyrightable.
In many cases, an artist may produce a work that is a mix of their own expression and AI-generated content. For example, a musician may use AI to make a song that uses AI-generated instrumentation and vocals but has the author’s original lyrics. In this case, the lyrics would be copyrightable, but the vocal and instrumentation output would not be. The Copyright Office states that an artist’s “own creative expression” will be protected by copyright, but not the AI-generated output itself (24). Additionally, an artist who arranges, selects, or coordinates AI-generated material may be able to claim copyright over the arrangement. The Copyright Office’s AI Copyright Registration Guidance details how artists seeking to register their creations should disclose AI usage.
Conclusion
As it stands currently, fully AI-generated music is not copyrightable. Artists using AI music software are not entitled to any copyright protection for outputs generated from their prompts. Notably, such output could even be infringing upon an original artist’s copyright or right of publicity protecting one’s exclusive right to control the commercial use of their identity if it is not considered fair use. This is an emerging issue highlighted by Ghostwriter977’s “Heart on My Sleeve,” which was an AI-generated song mimicking Drake and The Weeknd. Musicians looking to use AI in the creation of their music should be aware of these limitations around the technology’s use and copyrightability, and take care not to use the technology in a way that may be considered infringement and possibly leading to civil fines, injunctions against use and even criminal penalties in worst cases.