RIAA Quickly Fires Back Against Suno CEO’s ‘Transformative’ Comments As Generative AI Training Models Take Center Stage

  • Save

  • Save
Moments after the major labels filed copyright infringement actions against AI music platforms Suno and Udio, the former’s CEO engaged in a public back-and-forth with the RIAA. Photo Credit: Steve Johnson

Why confine legal battles to the courtroom? Immediately following news of the major labels’ massive copyright infringement lawsuits against Suno and Udio, the former AI music service’s CEO engaged in a testy war of words with the RIAA.

That back-and-forth, a testament to the disputes’ high-stakes nature and the broader significance of protected media’s use in generative AI, compelled the trade organization to put out an afternoon statement yesterday.

Beginning on the other side of the confrontation, Suno CEO Mikey Shulman in widely circulated comments defended his company’s technology as “transformative” and “designed to generate completely new outputs, not to memorize and regurgitate pre-existing content.”

Not stopping there, the exec reiterated his platform’s intentional lack of support for text-to-music “prompts that reference specific artists” and accused the plaintiff labels of reverting “to their old lawyer-led playbook” instead of engaging in “a good faith discussion.”

“Suno is built for new music, new uses, and new musicians,” concluded Shulman, with those “musicians” presumably referring to anyone capable of typing a single-sentence prompt. “We prize originality.”

According to the RIAA, Shulman in the lengthy response avoided the main question concerning his platform’s alleged ingestion of protected media.

“Suno continues to dodge the basic question: what sound recordings have they illegally copied?” the RIAA asked in the quickly distributed follow-up.

“In an apparent attempt to deceive working artists, rightsholders, and the media about its technology,” the organization proceeded, “Suno refuses to address the fact that its service has literally been caught on tape – as part of the evidence in this case – doing what Mr. Shulman says his company doesn’t do: memorizing and regurgitating the art made by humans.

“Winners of the streaming era worked cooperatively with artists and rightsholders to properly license music. The losers did exactly what Suno and Udio are doing now,” the entity concluded.

Though it perhaps goes without saying, there’s quite a lot riding on the central issue of whether training generative AI systems on protected media – the “caught on tape” line in the RIAA’s statement seemingly refers to related remarks from a Suno exec – is transformative and constitutes fair use.

While the obvious answer is a resounding “no” – among other things, developers would have trained their models solely on public-domain works if doing so was viable – many in the AI community are of the opposite stance. That includes Anthropic CEO Dario Amodei, who in April doubled down on the belief as his company fends off a separate infringement suit.

Time will reveal whether litigation can afford rightsholders their due compensation from generative AI giants and, more pressingly, bring about much-needed changes for future training practices as well as adjacent recordkeeping. However, even those solutions won’t halt the unprecedented technology’s sweeping impact, which could well undermine human creativity and a whole lot else in the long term.