A Voice Apart From the Song: Why Artists Are Treating Their Voice as an Asset

What it means for a voice to be an asset

In rights terms, an asset is something that can be owned, licensed and valued on its own. A voice has seldom been described that way because it almost always arrives attached to something else. Loredana Cacciotti, a music lawyer and founder of Future Play Music, noted that people connect it to a sound recording, an email, a video or a person, and rarely discuss it as having commercial value in itself.

Devin Kennedy, an independent artist, singer-songwriter and producer, said the question is new to him. Unless he was putting his voice on someone else’s song or collaborating, he said, he would not have thought about it until the last couple of years. He had not viewed it as “its own standalone asset until these conversations have kind of come up in the last year or so.

What changed when AI arrived

A founder of the New York entertainment and media law firm Weissman Law - Adam Weissman described the before and after. Before AI, he said, “this was never really an active concern for anybody.” When a voice was replicated, it was treated as a one-off, an anomaly or a joke, because doing it convincingly took a skilled editor or a very good impersonator.

Since AI tools reached the general public, Weissman said, he has seen people’s voices, and their likenesses, replicated in commercials for brands they have no affiliation with. He described it as happening with increasing frequency because it is now so easy. Anyone who makes a living publishing songs, videos or photos of themselves has supplied the raw material for a clone that can be made to say or do anything. The issue, he said, has gone from something nobody considered a problem “to something that’s top of mind for a ton of the people we work with at this point.”

Where artists draw the line

Kennedy described himself as “a little bit of a music recording purist at heart,” which puts him on the side that is harder to convince to license a voice or give it away easily. He separated an open-ended arrangement from a specific use. A clone license that is open-ended, or granted on “one signature,” left him “very skeptical.”

Some uses looked more attractive. One was localization into other languages, so that, in his words, “I’m able to reach people in other countries with the same music, not have to learn ten different languages to do that.” Another was iterative versions of music on streaming platforms. His openness depended on the use being the right one, and on rights holders being protected and compensated.

Two developments made him “really excited about protecting” a voice as a standalone asset: Taylor Swift’s effort to create legal precedent around her likeness, and Tennessee’s ELVIS Act, which added voice to the state’s right-of-publicity law and took effect in July 2024.

Why the distinction matters now

Because a cloned voice can exist without any specific recording or composition, the rights that music contracts were built around do not obviously reach it. Cacciotti’s opening question to Kennedy put the practical issue plainly: how comfortable is an artist with having a voice licensed separately from the rest of their music? His answer was conditional. He was skeptical of open-ended licenses and open to specific uses.

The infrastructure question runs in parallel. Dae Bogan, head of third-party partnerships at The Mechanical Licensing Collective (The MLC), said his organization represents musical works, meaning lyrics and melodies, “but we don’t represent voices.” Systems built to track and pay for recordings and compositions were not designed around a voice that exists apart from both.