The ELVIS Act: Protecting the Music Industry from Artificial Intelligence
The ELVIS Act: Protecting the Music Industry from Artificial Intelligence
Written by Savannah Yamat
Artificial Intelligence (“AI”) is rapidly revolutionizing virtually every industry, and the world of music is no exception. AI has revolutionized the music industry by providing tools artists can use to compose, perform, and produce music. However, this technological evolution comes with both exciting opportunities and significant challenges. This includes challenges to intellectual property rights, such as traditional copyright and trademark protections. AI technology disrupts traditional models of music creation, distribution, and ownership. As this technology grows and develops, so must our legal system. Although Congress has not yet enacted a broad federal law for AI development or use, state legislatures have made efforts to regulate AI usage.
In this post, I will be discussing the implications of AI in the music industry by focusing on a recent piece of legislation passed in Tennessee: The Ensuring Likeness and Image Security (“ELVIS”) Act[1]. This law aims to protect musicians from the unauthorized use of their voices through AI technologies, including audio deepfakes and voice cloning. This post will explore how the ELVIS Act could set a precedent for regulating AI in the entertainment industry, offering both a solution to protect artist legacies and a framework for tackling new copyright and trademark challenges in an era of AI-driven creativity.
- IP Challenges in the Age of AI
The use of AI in music production raises many significant questions. One of the most significant is how we can reconcile the capabilities of AI systems with the legal frameworks that govern intellectual property.
Copyright and trademark law play an essential role in protecting the creativity, livelihood, and identity of musicians and music creators. Copyright law protects original works of authorship, including a song’s underlying composition or lyrics.[2] It also protects sound recordings.[3] The copyright owner is given the exclusive right to reproduce the work, distribute it, and perform it publicly. Trademarks protect the branding and identity of artists, bands, record labels, and other music-related products and services.[4] Examples include band names, a band or artist’s logo, album names, and tour names.[5] However, traditional copyright and trademark rights may be violated with the use of AI. Many AI music-generation systems are designed using a type of machine learning called deep learning.[6] In doing so, AI systems are trained by exposure to large datasets of music.[7] The AI learns harmony, rhythm, instrumentation, and melody patterns, which it can then use to generate new works. This complicates traditional copyright laws which grant artists the exclusive right to distribute and perform their work(s). For example, if an artist wants to “sample” another artist’s song or lyrics, they will typically have to get clearance from the original artist or the record label that owns the musical catalog.[8] Similarly, AI, in a way, “samples” existing works of art: it composes new pieces based on the dataset(s) it has been trained on. Thus, a serious question arises: If an AI system is trained on copyrighted music and produces a new work based on that material, has it used copyright-protected content without permission—and if so, who is legally responsible?
Another challenge regarding AI music creation is ownership. In traditional music creation, the person who creates, performs, or composes a piece of music holds copyright. However, when using AI to generate musical pieces, authorship is not so easily established. Under current legal framework, only human creators can be awarded copyright protection for their works. For example, the U.S. Copyright Office has a human requirement for granting a copyright in a work of art: “The U.S. Copyright Office will register an original work of authorship, provided that the work was created by a human being.”[9] This creates a dilemma for AI-generated music, as it challenges the notion of authorship itself. For an AI-generated song, who is the author? The individual who prompted the AI system? The AI system itself? Or is it the author(s) of the pieces of work the AI system has been trained with? Currently, there is no legal precedent that answers this question. The issue of ownership remains ambiguous.
2. The Growing Role of AI in the Music Industry
The first computer-generated piece of music—the Illiac Suite—was created in 1956 through the computing power of the ILLIAC (Illinois Automated Computer).[10] The Illiac Suite is a four movement composition for a string quartet. However, AI usage in music has advanced much farther than the Illiac Suite. The use of AI in the music industry is growing rapidly and being used in many contexts: from music creation to distribution of marketing.
AI has utility in nearly all stages of music production. Tools such as ReMi[11], Claude Sonnet 3.5[12], and ChatGPT[13] are able to provide assistance with the lyric writing process. Softwares such as eMastered[14], iZotope Ozone[15], and LANDR[16] aid in the mixing and mastering process. AI programs such as AIVA[17], Juke Box[18], and MuseNet[19] are capable of generating entirely new songs, including lyrics.
In addition to music production, AI is being used in the areas of music discovery and recommendation. Streaming platforms like Spotify, Apple Music, and YouTube Music all use AI-driven algorithms to recommend songs for users based on user data, listening history, likes, search queries, and track skips[20]. Spotify even introduced an AI DJ which serves a personal guide—complete with an AI-generated voice—to personalize the music listening experience for users.[21]
One of the most controversial aspects of AI in music generation is AI’s ability to replicate human voices, also known as voice cloning. AI is able to create highly accurate imitations of a person’s voice. A quick YouTube search will yield thousands of results of AI covers (e.g., a 1960s Jazz Motown Soul cover of 2Pac’s “Dear Mama”[22] or a Hair Metal cover of the Spice Girl’s “Wannabe”[23]). Although these cover songs appear harmless and the ability to clone someone’s voice may have broad utility for the entertainment industry, AI voice imitation poses many legal and IP issues. Although voices cannot be trademarked[24], copyrighted or copyrighted, the impersonation of someone’s voice could still very likely be considered infringement or a violation of an individual’s right to publicity (if used in a commercial setting). However, there is a lack of legal protection against AI voice cloning. The “ELVIS” Act was designed to help fill this void.
3. The “ELVIS” Act: A Potential Solution
The state of Tennessee has long been synonymous with music, with its capital, Nashville, earning the title “Music City” due to its deep historical, cultural, and artistic ties to the music industry.[25] With this background, it is fitting that Tennessee has become the first state to introduce legislation combatting unauthorized AI impersonation. The Ensuring Likeness and Image Security (“ELVIS”) Act directly addresses the growing concerns surrounding the use of AI in the music industry; specifically, the potential for AI to create music that mimics the style or likeness of established artists without consent.[26]
The ELVIS was signed on March 21, 2024 and went into effect on July 1, 2024. In a unanimous, bipartisan vote, the bill passed both the Tennessee House and Senate, with 93 ayes and 0 noes in the House, and 30 ayes and 0 nays in the Senate.[27] The ELVIS Act was not a standalone bill, but an amendment to Tennessee’s Personal Rights Protection Act of 1984.[28] The Tennessee Personal Rights Protection Act was the result of litigation undertaken by the estate of Elvis Presley to gain control of how his likeness could be used after his death.[29] The Personal Rights Protection Act created a property right in an individual’s name, photograph, and likeness.[30] These rights do not expire upon death and are descendible to an individual’s executors, assigns, heirs, or devisees.[31]
The ELVIS Act builds on these regulations and expands the individual’s control over their likeness by adding protections to the individual’s sound and voice.[32] The ELVIS act directly addresses deepfakes and voice clones; it prohibits people from using AI to mimic a person’s voice without their permission.[33] The act gives the individual the fundamental right to control the commercial use of their identity—specifically, their name, image, likeness, and voice.[34] The ELVIS Act makes it unlawful for someone to exploit an individual’s identity for commercial gain without consent—including use in advertisements, endorsements, and media misrepresentations.[35] Any nonconsensual use of someone’s image, voice, or persona in AI-generated content can be criminally enforced as a Class A misdemeanor punishable by up to 11 months and 29 days of incarceration and fines reaching $2,500.[36] Additionally, the law grants individuals the right to file a civil lawsuit against anyone who infringes upon its regulations.[37]
Like any other piece of legislation, the ELVIS Act has its strengths and weaknesses. Tennessee Senate Majority Leader Jack Johnson described the ELVIS Act as “first-of-its-kind,” and it truly is.[38] The ELVIS Act is the first piece of legislation directly designed to combat AI-generated imitation or impersonation. Its greatest strength is how it grants musicians, songwriters, and creatives ultimate control over their works of art and protects them from unauthorized use and exploitation for commercial gain. It also empowers them to challenge unauthorized use and seek damages. The ELVIS Act reinforces that authorship belongs to the humans who create the works of art and bridges some of the gaps that exist in modern copyright and trademark law. However, there are also some negative aspects that must be explored. AI is a tool. Narrowing the context in which AI can be used and regulating it can be seen as a form of suppression of creativity and free expression. Furthermore, enforcement of the act may be difficult in multiple ways. Parody and satire have been long recognized as protected forms of expression under the First Amendment. Can individuals claim the First Amendment Right of Speech to justify AI-generated imitation or impersonation? Also, many AI tools are decentralized and/or used anonymously. Many AI programs are able to run on personal or mobile devices, creating difficulty when tracing the origin of infringing works. It is highly likely that many unauthorized uses will go undetected and unpunished.
Despite its shortcomings, the ELVIS Act is a pivotal first step in regulating AI. It sets a precedent for other states and, potentially, for the federal government to consider when addressing similar issues. It has a great potential for utility, but its long-term effectiveness will depend on its ability to be properly enforced, and its ability to adapt.
4. Conclusion
In conclusion, the ELVIS Act marks a significant step forward in addressing the intersection of AI and IP rights in the music industry. The new protections it offers against AI-generated impersonation sets an important precedent for federal and state legislatures. While the ELVIS Act is the “first-of-its-kind,” other state legislatures have begun to tackle the intersection of law and AI. California’s Governor Newsom recently signed legislation regulating deepfakes.[39] While these state-level efforts are important first steps, a unified and thorough approach is pivotal to regulate AI’s growing influence on industries like music, entertainment, and beyond.
[1] Tenn. Gen. Assembly, HB 2091, 113th Gen. Assembly (2023), https://www.capitol.tn.gov/Bills/113/Bill/HB2091.pdf
[2] U.S. Copyright Office, What Musicians Should Know About Copyright, https://www.copyright.gov/engage/musicians/
[3] Id.
[4] Jamie Brehm, Trademark vs. Copyright: Key Differences for Music Licensing, STOCKMUSIC.NET. (Feb 23, 2022), https://stockmusic.net/blog/trademark-vs-copyright-key-differences-for-music-licensing/
[5] World Intellectual Property Organization (WIPO), Trademarks in Music, https://www.wipo.int/en/web/music/trademarks
[6] Jean-Pierre Briot, Gaëtan Hadjeres, and Franc¸ois-David Pachet, Deep Learning Techniques for Music Generation – A Survey, arXiv (August 7, 2019), ehttps://arxiv.org/pdf/1709.01620
[7] Id.
[8] Taylar-Simone McCants, Obtaining Permission Before Sampling Music, FINDLAW (May 22, 2024), https://www.findlaw.com/smallbusiness/intellectual-property/obtaining-permission-before-sampling-music.html
[9] U.S. Copyright Office, Copyrightable Authorship: What Can Be Registered, COPYRIGHT.GOV, (January 28, 2021), https://www.copyright.gov/comp3/chap300/ch300-copyrightable-authorship.pdf
[10] Ayal Zipris, Iliac Suite, ILLINOIS DISTRIBUTED MEDIUM, https://distributedmuseum.illinois.edu/exhibit/illiac-suite/ ; alex di nunzio, Iliac Suite for String Quartet [1/4], YOUTUBE (December 4, 2011), https://www.youtube.com/watch?v=n0njBFLQSk8&embeds_referring_euri=https%3A%2F%2Fdistributedmuseum.illinois.edu%2F&source_ve_path=Mjg2NjQsOTY3MTQ
[11] Suno, What is ReMi?, https://help.suno.com/en/articles/3599681
[12] Anthropic, Introducing Claude 3.5 Sonnet, ANTHROPIC (June 20, 2024), https://www.anthropic.com/news/claude-3-5-sonnet
[13] OpenAI, ChatGPT, https://chatgpt.com/
[14] eMastered, https://emastered.com/
[15] iZotope, Ozone 12, IZOTOPE, https://www.izotope.com/en/products/ozone.html?srsltid=AfmBOoqRmdN2NuS77xVzEGDmDBCbjuwShKlI0SHFRNsp5sqdvEYhVW4k
[16] LANDR, https://www.landr.com/?srsltid=AfmBOoqjylUG_8KN4V3XlWgNcgbPRlaqZe8tLPv1T_2qhaulmBL_UvTP
[17] AIVA, https://www.aiva.ai/
[18] OpenAI, Jukebox, https://openai.com/index/jukebox/
[19] OpenAI, MuseNet, https://openai.com/index/musenet/
[20] Spotify, Understanding Recommendations on Spotify, Safety and Privacy Center (last updated 2025), https://www.spotify.com/us/safetyandprivacy/understanding-recommendations
[21] Spotify, Spotify Debuts a New AI DJ, Right in Your Pocket, SPOTIFY NEWSROOM (February 22, 2023), https://newsroom.spotify.com/2023-02-22/spotify-debuts-a-new-ai-dj-right-in-your-pocket/
[22] Re Soul’d, Dear Mama – 2Pac (1960s Jazz Motown Soul AI Cover), YOUTUBE (October 17, 2025), https://www.youtube.com/watch?v=hxwYzNDgvfU&list=RDhxwYzNDgvfU&start_radio=1
[23] Hair Metal Covers, Can WANNABE be good as Hair Metal?, YOUTUBE (November 174, 2025), https://www.youtube.com/watch?v=ZmnyzXhYF7M&list=RDZmnyzXhYF7M&start_radio=1
[24] Ivy Attenborough, Voices, Copyrighting, and Deepfakes, IP WATCHDOG (October 14, 2020), https://ipwatchdog.com/2020/10/14/voices-copyrighting-deepfakes/
[25] Wikipedia, Nashville, Tennessee, WIKIPEDIA, https://en.wikipedia.org/wiki/Nashville,_Tennessee
[26] Tenn. Gen. Assembly, HB 2091
[27] Tennessee Government Access, House Floor Session – 44th Legislative Day, GRANICUS (March 4, 2024) https://tnga.granicus.com/player/clip/29778?view_id=703&meta_id=803640&redirect=true
[28] Personal Rights Protection Act, Tenn. Code. Ann. § 47-25-1101 et seq. (1984). https://preslaw.info/wp-content/uploads/2017/02/prpa.pdf
[29] State ex rel. Elvis Presley Int’l Mem’l Found. V. Crowell, 733 S.W.2d 89 (Tenn. Ct. App. 1987). https://law.justia.com/cases/tennessee/court-of-appeals/1987/733-s-w-2d-89-2.html
[30] Tenn. Code. Ann. § 47-25-1101 et seq. https://preslaw.info/wp-content/uploads/2017/02/prpa.pdf
[31] Id.
[32] Tenn. Gen. Assembly, HB 2091
[33] Id.
[34] Id.
[35] Id.
[36] Id.
[37] Id.
[38] Tennessee Office of the Governor, Photos: Gov. Lee Signs ELVIS Act Into Law (March 21, 2024), https://www.tn.gov/governor/news/2024/3/21/photos–gov–lee-signs-elvis-act-into-law.html
[39] Governor Gavin Newsome, Governor Newsom signs bills to combat deepfake election content (September 17, 2024), https://www.gov.ca.gov/2024/09/17/governor-newsom-signs-bills-to-combat-deepfake-election-content/