Robots in the Recording Studio: IP Rights in the AI Music Revolution
As artificial intelligence increasingly expands into the music industry, important questions have emerged around ownership, royalties, and the proper scope of intellectual property rights for AI-generated works. Similar to the wave of digitization that transformed music distribution through streaming platforms starting in the 2000s, the current rise of AI has the potential to bring about both new opportunities and new legal and ethical issues over intellectual property claims.[1] This post broadly explores intellectual property (IP) rights regarding AI music platforms and considers how our legal framework balances the transformative benefits this new technology brings while protecting artist’s rights.
AI music platforms come in various forms. Some create entirely new compositions in seconds, others generate instrumentals based on given prompts, and a few even clone voices.[2][3] These AI generations are often indistinguishable from human compositions and are trained on large, unlicensed datasets gathered from the internet.[4] Tools like Suno or Beatblender are illustrations of this type of AI-generated music production.[5]
Alternatively, other platforms foster an AI-assisted approach to music production, where AI is a complementary rather than primary tool.[6] In this form, rather than generating entire songs or beats from a prompt, AI refines the creator’s vision through handling limited production tasks, providing inspiration, or offering suggestions while the artist retains broad control over the creative process.[7]
Because of their efficiency, both of these types of platforms often reduce production costs by automating tasks that once required hours of manual labor, like music inspiration, mixing and mastering, or song restoration.[8] Music production is a resource-intensive, time-consuming process, and AI tools are lowering that burden, perhaps helping “democratize music making,” allowing individuals with less resources to compete with big music industry players.[9]
On the flipside, the ease of generating music has led to widespread legal and ethical concerns. Given that platforms are trained on and draw from large, unlicensed datasets gathered from the internet, AI uses and especially fully AI-generated songs are largely just a computer algorithm, rendering assignment of copyright ownership difficult.[10] Furthermore, human authorship is required under current U.S. copyright policy, creating a line-drawing problem regarding how much AI can be used and at what point a work turns from a human-authored song that our legal IP system is equipped to handle to a machine-generated work drawing on countless datasets of other artists’ expressions.[11]
Moreover, AI developers may include terms of service that limit or expand user rights. For example, some agreements might require users to grant the platform a perpetual license to any generated content, effectively transferring partial ownership to the platform itself; one example of such terms is Suno.[12] Suno’s terms effectively state that by uploading or creating content through Suno’s service, individuals give Suno (and associated parties) broad permission to use and share that content and any derivatives of it indefinitely and without pay.[13] These types of terms can lead to massive royalty disputes, and if the platforms claim partial ownership of the compositions can lead to artists’ creative control and earnings being undermined.[14]
Additional concerns arise out of the datasets used to train AI music platforms. First, if the AI has been trained on copyrighted tracks without proper authorization, the output could be deemed an unauthorized derivative work or could infringe on the rights of the original creators.[15] This seems to be the case based on the recent ruling of Thomson Reuters Enterprise Centre GMBH v. Ross Intelligence Inc. which “struck down the fair use defense in the context of AI training data.”[16] Furthermore, concerns arise out of how the datasets that AI music platforms rely on are selected. As Adam Clair puts, “whose voices are included in that huge crate of music, and whose are left out?” Selective, biased, and predictable models run the risk of excluding and forcing out underrepresented cultures and languages from the dataset.[17]
Voice cloning tools introduce potential right of publicity claims if used without the artist’s consent.[18] While it is debated whether existing right of publicity laws prevent AI voice cloning tools, states like Tennessee are amending their laws or enacting new ones to address AI content, with the ELVIS Act of 2024 as one example.[19]
In response to growing concerns over AI generated art, the U.S. Copyright Office introduced policies that delineate the specific registration criteria for AI-generated works, such as requiring a substantial human contribution.[20] Additionally, the policies noted that mere selection of prompts does not itself make a work eligible for copyright protection—only human contributions are potentially copyrightable when a work includes both human and AI content, and the use of AI as a tool to enhance the human creative process does not render the entire work uncopyrightable.[21] Time will tell whether these policies will strike the best balance of protecting artists’ rights while allowing the benefits and innovation of AI to continue.
Looking ahead, we may see test cases that provide a clearer roadmap for how courts distinguish between human-authored work and AI-generated content. Artists and labels may continue to tinker with contract clauses to manage various risks until potential broader legislative action is taken. Ultimately, AI music platforms offer the potential to redefine how we create and enjoy music. But without proactive steps to clarify intellectual property rights—ranging from copyright ownership to voice likeness claims—both innovation and the livelihoods of artists may be at risk. The key will be finding a delicate balance that respects creativity, incentivizes new technologies, and ensures fair compensation in the evolving AI-driven music landscape.
[1] Kyle Wiggers, Amanda Silberling, AI music generators could be a boon for artists — but also problematic (Oct. 7, 2022), https://techcrunch.com/2022/10/07/ai-music-generator-dance-diffusion/.
[2] Alberto Martinez Jr, The Rise of AI-Generated Music: What It Means for Artists (Jan. 4, 2024), https://flourishprosper.net/music-resources/the-rise-of-ai-generated-music-what-it-means-for-artists/
[3] Elizabeth Shields, The AI Doppelgänger Dilemma: Cloned Voices in the Music Industry, 48 SEATTLE U. L. REV. ONLINE 71 (2025), https://digitalcommons.law.seattleu.edu/sulr_supra/32/
[4] Virginie Berger, AI’s Impact On Music In 2025: Licensing, Creativity And Industry Survival (Dec. 30, 2024), https://www.forbes.com/sites/virginieberger/2024/12/30/ais-impact-on-music-in-2025-licensing-creativity-and-industry-survival/
[5] Jocelyn Johnson, What is AI in Music Production? (Sep. 18, 2024), https://www.trackclub.com/resources/aiinmusicproduction.
[6] Id.
[7] Id.
[8] Id.
[9] Amanda Hoover, AI-Generated Music Is About to Flood Streaming Platforms (Apr. 17, 2023), https://www.wired.com/story/ai-generated-music-streaming-services-copyright/
[10] Virginie Berger, AI’s Impact On Music In 2025: Licensing, Creativity And Industry Survival (Dec. 30, 2024), https://www.forbes.com/sites/virginieberger/2024/12/30/ais-impact-on-music-in-2025-licensing-creativity-and-industry-survival/
[11] U.S. Copyright Office Policy on AI Authorship, U.S. Copyright Office, https://www.copyright.gov/
[12] Suno Terms of Service, (Jun. 30, 2024), https://suno.com/terms/
[13] Id.
[14] The Implications of AI Music in the Music Industry (Jan. 28, 2025). https://mubert.com/blog/how-ai-is-changing-music-licensing-legal-battles-copyright-challenges-and-the-future-of-the-industry.
[15] Amanda Hoover, AI-Generated Music Is About to Flood Streaming Platforms (Apr. 17, 2023), https://www.wired.com/story/ai-generated-music-streaming-services-copyright/
[16] Monique N. Bhargava Mitesh P. Patel Katherine Litaker, Court shuts down AI fair use argument in Thomson Reuters Enterprise Centre GMBH v. Ross Intelligence Inc. (Mar. 3, 2025). https://www.reedsmith.com/en/perspectives/2025/03/court-ai-fair-use-thomson-reuters-enterprise-gmbh-ross-intelligence.
[17] Adam Clair, What AI in music can — and can’t — do (Aug. 5, 2024). https://www.vox.com/the-highlight/358201/how-does-ai-music-work-benefits-creativity-production-spotify.
[18] Myriah V. Jaworski, Chirag H. Patel, ROP on the Rise: Right of Publicity Claims Will Rise as States Address AI Generated Deepfakes and Voice Cloning (Apr. 15, 2024), https://www.clarkhill.com/news-events/news/rop-on-the-rise-right-of-publicity-claims-will-rise-as-states-address-ai-generated-deepfakes-and-voice-cloning/
[19] Id.
[20] M. Oren Epstein, Stuart D. Levi, Jordan Feirman, Mana Ghaemmaghami, MacKinzie M. Neal, Copyright Office Publishes Report on Copyrightability of AI-Generated Materials (Feb. 4, 2025), https://www.skadden.com/insights/publications/2025/02/copyright-office-publishes-report.
[21] M. Oren Epstein, Stuart D. Levi, Jordan Feirman, Mana Ghaemmaghami, MacKinzie M. Neal, Copyright Office Publishes Report on Copyrightability of AI-Generated Materials (Feb. 4, 2025), https://www.skadden.com/insights/publications/2025/02/copyright-office-publishes-report.