You can imitate a musician’s style, at least to a point. It’s the sincerest form of flattery, and we’re all guilty of some version of it, ranging from just for fun at bar karaoke to professional cover bands. But when AI takes the things that make a performer sound like themselves and turns them into a commercial product, that seems like a different story. That argument is the center of a new lawsuit brought by Jason Isbell and other musicians.
If you’re a fan of Drive-By Truckers or Isbell’s solo stuff, you’ll recognize the six-time Grammy-winning Alabama singer-songwriter and guitarist. Joined by David Lowery of Camper Van Beethoven and Cracker, Texas blues musician Guy Forsyth, and Grammy-winning saxophonist Eduardo “Ed” Calle, Isbell says Cambridge-based AI music company Suno crossed a line. In a recent putative class action, the musicians accuse Suno of using their identities without consent to generate artist-evoking songs, descriptions, and images when users enter a musician’s name.
What Does Suno Do?
Suno lets users generate music from text prompts. Users can type an instruction, add lyrics, upload audio, or request an instrumental track. Paid users can access additional tools, including audio-input features and the ability to download and commercially use generated music.
According to the complaint, Suno went beyond being a general-purpose music generator. The musicians claim Suno built a platform that uses a performer’s name as a shortcut to retrieve and deploy that performer’s identifying characteristics.
The basic allegation is this: Type in a famous musician’s name, and Suno allegedly produces a whole package designed to call that person to mind. That can include a newly generated song, a description of the song’s musical attributes, and an image. The plaintiffs say this is not just an AI system generating “country music” or “blues music.” They argue that it is a commercial tool built to evoke particular people.
After the suit was filed, Suno said the claims lack merit and that it would defend the case. The company also said it blocks prompts for specific artists’ names and copyrighted songs. But the complaint claims the filters are inconsistent and easy to evade.
A Name Game
The complaint includes a lengthy list of Suno demonstrations. The plaintiffs say that entering musicians’ names (sometimes with no other prompt at all) caused Suno to create outputs that evoked those musicians’ styles, instruments, vocal qualities, histories, or visual associations.
For example, the lawsuit alleges that entering “Ed Calle” produced Latin-jazz music featuring saxophone flourishes and images of a saxophone beneath palm trees. Typing “Guy Forsyth,” the complaint says, generated a blues-rock track associated with a smoky live-performance setting.
The lawsuit also points to demonstrations involving well-known artists outside the proposed class, including Buddy Guy, Carly Simon, Mavis Staples, The Chicks, Israel Kamakawiwo’ole, Taylor Swift, Tom Waits, Michael Jackson, Snoop Dogg, Bad Bunny, and Chief Keef.
In some cases, the complaint claims Suno accepted names directly. In others, it alleges that inserting spaces between the letters of a name (e.g., “t a y l o r s w i f t”) worked around the company’s supposed safeguards. The plaintiffs say these examples show that Suno’s system recognizes a performer’s name as more than a collection of words. In their view, the name works as a key that calls up a set of artist-specific features embedded in the AI model.
That is a much different proposition than simply using an artist’s name to help listeners locate that artist’s authorized work on a streaming service.
The ‘Borg’ Theory
The complaint opens with a surprisingly sci-fi-heavy comparison: the Borg from Star Trek.
For the uninitiated, the Borg are a collective of cybernetic beings that assimilate other people and civilizations, absorbing their distinctive qualities into a shared machine intelligence. The complaint argues that Suno does something analogous with musicians’ identities: it allegedly takes the human characteristics that make a performer recognizable, converts them into machine-readable data, and deploys them as part of a product that generates new music on demand.
It might not read like your typical legal brief, but the metaphor makes the plaintiffs’ theory clear. They claim Suno did not need to reproduce a particular Jason Isbell song to exploit Jason Isbell’s identity. In their view, the company extracted the elements that make listeners say, “That sounds like Jason Isbell,” and made those elements available through its platform.
Suno has not yet filed a response to the complaint, so these are allegations that will need to be tested in court.
Not Really a Copyright Case
The lawsuit goes out of its way to distinguish its claims from the many AI-related copyright fights already making their way through the courts. The plaintiffs are not claiming ownership of a musical genre. They say they are not trying to stop people (or even AI systems) from making country, blues, folk, hip-hop, jazz, or any other kind of music. And they do not say their claims depend on Suno reproducing a particular recording.
Instead, they rely on rights-of-publicity laws. Those laws generally protect people from unauthorized commercial use of their name, image, likeness, voice, or other identifying aspects of their persona. The complaint alleges violations of Illinois’ Right of Publicity Act, New York’s right-of-publicity statutes, California’s common-law right of publicity, and similar laws in several other states.
The plaintiffs also allege that Suno created and stored “voiceprints” (numeric representations of vocal characteristics) without the consent allegedly required under Illinois’ Biometric Information Privacy Act.
What Happens Next?
The legal theory is ambitious. It asks the court to recognize that an AI model can commercially use a musician’s identity even if it does not use a photo, a direct voice clone, or an exact recording. The plaintiffs seek damages, an order stopping the alleged conduct, and a declaration that Suno violated their publicity and biometric-privacy rights.
No court has yet decided whether the allegations are true, whether Suno’s technology uses musicians’ identities in the manner alleged, or whether those state-law protections apply to AI training and generated music. Suno will have the chance to respond, seek dismissal, and challenge whether the case can proceed as a class action.
However the case turns out, it raises a question that reaches far beyond music: When an AI product can evoke a recognizable person on demand, where does inspiration end and commercial exploitation begin?
