What do Dolly Parton, Miley Cyrus, Aretha Franklin and Justin Timberlake have in common? These famous singers were all born in Tennessee – the capital of which, Nashville, is the biggest music recording and production centre in the United States outside of New York City. Other famous musicians who relocated to the “Song writing Capital of the World” to launch their careers include Taylor Swift and Johnny Cash. Tennessee’s music industry supports more than 61,000 jobs across the state, and contributes a total economic impact of $15.6 (£12.5) billion to the region’s economy.
It's no surprise then, that Tennessee would lead the way in protecting musician’s voices. Against a backdrop of rapidly advancing AI technologies which make it possible to mimic a person's voice with alarming accuracy, Tennessee lawmakers have become the first in the world to pass a law prohibiting the use of AI-replicated voices without the source artist’s consent.
The Ensuring Likeness Voice and Image Security (“ELVIS”) Act of 2024 takes effect from 1 July, was endorsed by the Recording Academy (known for the Grammy’s), the Recording Industry Association of America, and SAG-AFTRA.
Updating a 40 year-old law for the age of AI
The ELVIS Act amends Tennessee’s existing statutory publicity rights set out in the Personal Rights Protection Act of 1984, which itself was the result of litigation surrounding how Elvis Presley’s image could be used after his death. You can see the redline of amendments, here.
It’s important to note that the Act does not mention artificial intelligence explicitly (although it does mention algorithms in 6(a)(c), discussed below). Accordingly, one could still break the law, even if they didn’t use AI per se. That said, it’s obvious that minimising threats posed by AI was the key priority when drafting this legislation. Although voice impersonation has been around for ages, new generative AI tools have made voice cloning easy, inexpensive, and incredibly realistic. Just listen to Heart On My Sleeve, a song that features AI-generated vocals cloned (without permission, it turns out) off of Drake and The Weeknd, or indeed Drake's own Taylor Made Freestyle, a diss-track against fellow rapper Kendrick Lamar which features an AI-cloned Tupac Shakur (likewise without permission).
These innovations have pushed lawmakers to now include “voice” in the protected characteristics of one’s persona (public image). Voice is defined in the Act as:
a sound in a medium that is readily identifiable and attributable to a particular individual, regardless of whether the sound contains the actual voice or a simulation of the voice of the individual.
Under Section 6(a)(2), a person can now be sued if they “publish, perform, distribute, transmit, or otherwise make available to the public an individual's voice or likeness, with knowledge that use of the voice or likeness was not authorized by the individual.”
Section 6(a)(3) also creates a new civil action for the supply of “an algorithm, software, tool, or other technology, service, or device” where the primary purpose or function “is the production of an individual's photograph, voice, or likeness without authorisation from the individual”. In other words, a person can be sued if they distribute or provide technology designed to capture or clone someone else’s likeness without consent.
Property vs. Human Right: A Normative Discussion?
As a practising lawyer and vocal advocate of strong publicity rights, there’s a lot to celebrate, here. But despite its forward-thinking provisions, the ELVIS Act leaves me wondering if this wasn’t a missed opportunity to transform the protection into a more substantial, inalienable human right.
The statute makes clear that “every individual has a property right in the use of that individual's name, photograph, voice, or likeness in any medium in any manner” (Section 4(a)). Framing the protection of one’s voice and likeness as a property right makes them inherently transferrable (for example by licence or assignment), and primarily serves economic interests. This approach is designed to allow the commercial transactions that underpin the music and entertainment industry, and aligns with the “labour theory” of copyright more generally.
However, this approach fails to capture the deeply personal and intrinsic value of a voice. One's voice goes beyond commercial interests or financial gain: "voice" is eponymous with opinion and expression for a reason.
I appreciate that a property right offers flexibility and ease when dealing with commercial contracts – and these are crucial when licensing songs between singers, record labels, radio stations, and beyond. But there’s also a valid argument for establishing a more robust and enduring protection under a human rights framework; one in which agency, autonomy and dignity are prioritised. This is especially true if AI versions of our personas – of our voices, likenesses, and physical beings – can be duplicated, modified, and distributed with just the click of a button.
Until then, the ELVIS Act is indeed good progress. I suppose that for now, it’s up to those in the entertainment business – to include the lawyers negotiating contracts – to do what they can to protect a musician’s voice from AI replication. Or, at the very least, to ensure the artists are paid fairly for it.