Record Labels License Music to AI Platforms They Once Sued

From courtroom to partnership

The music industry's relationship with AI took a sharp turn in 2024 when Universal, Sony and Warner sued AI music generators Suno and Udio for copyright infringement. Two years later, several of those same labels are licensing their catalogs to the platforms they accused of stealing their artists' work.

Suno, now valued at $5.4 billion, released its first models built with licensed music from Warner Music Group, BMG and Believe earlier this month. Rival Udio has struck licensing deals with Universal and Warner. The settlements began in 2025, with Universal partnering with Udio in October and Warner signing with both platforms in November.

The companies frame these arrangements as new revenue streams for artists. But critical details remain undisclosed: what artists will be paid, how many have opted in, and how much control individual musicians have over whether their work feeds AI training data.

Who controls the choice

The deals share a common structure. Artists signed to participating labels can theoretically opt in to having their music used and will receive credit and payment. But the reality is more complex.

Songs typically carry two separate copyrights—one for the composition (held by songwriters and publishers) and one for the sound recording (usually owned by the label). How AI products handle music where these rights are split remains unclear, and every contract differs.

"These companies are able to make deals on behalf of their artists, and their control over their music is pretty significant based on what major label contracts typically say," entertainment attorney Jane Davidson told the Los Angeles Times, which first reported these details.

Jack Brody, Suno's chief product officer, acknowledged that labels "ultimately have their contracts in place that will have some say in what an artist can and can't do." Whether opting out prevents an artist's recordings from training AI models or merely keeps them out of consumer-facing features also remains ambiguous.

Musicians push back

The American Federation of Musicians sued Universal and Warner in June, alleging the labels licensed recordings featuring union members to Suno and Udio without compensating those musicians or disclosing which recordings were involved. The labels have moved to dismiss the case.

Session musician Matt Evans, who has played brass on "The Late Late Show With James Corden" and other productions, described the situation as "a creepy and ominous cloud." His concern centers on compensation: "The fact that I could do something once and then it could be reused, and if that's not negotiated in a contract, then I never got to make a choice to receive additional compensation."

A proposed class action filed last month by Jason Isbell and other musicians alleges Suno allows users to generate songs evoking specific artists' identities without consent—a right-of-publicity claim the plaintiffs argue belongs to performers regardless of who owns their recordings.

Songwriter Michelle Lewis, co-founder of Songwriters of North America, noted the composition side of copyright is harder to defend: "With AI, it's so much easier to tell if you've infringed on the sound recording. You know what Beyoncé and Drake's voices sound like. But songwriting is so much more nebulous, so much harder to sue."

Litigation continues

Not all labels are settling. Sony filed a second lawsuit against Udio over more than 30,000 recordings and continues to sue Suno alongside Universal—which is simultaneously licensing music to Udio and Spotify. Sony and Universal filed a new lawsuit last week alleging Suno's latest model still infringes their copyrighted works.

Why it matters

These licensing deals represent a pivotal moment for how AI companies access training data and how creative industries balance technological adoption with artist compensation. The undisclosed payment terms and unclear consent mechanisms echo long-standing complaints about streaming royalties, but with potentially greater stakes. If AI tools can replicate the creative contributions of session musicians, composers and performers without transparent compensation structures, the deals could fundamentally reshape labor markets in music production. The legal battles still underway will likely determine whether artists gain meaningful control over their work in AI systems or whether labels can unilaterally license recordings for uses never contemplated in existing contracts.

The Los Angeles Times first reported these developments.