• by Wendy Davis  @wendyndavis, June 24, 2024

Major record labels on Monday sued the artificial intelligence companies Suno and Udio for allegedly using copyrighted material to train services that generate songs in response to requests by consumers.

“Foundational principles of copyright law dictate that copying protected sound recordings for the purpose of developing an AI product requires permission from rightsholders,” Sony Music, Universal Music Group and Warner Records write in complaints filed against the companies. The case against Udio was filed in federal court in New York, and the complaint against Suno was brought in federal court in Massachusetts.

The record labels claim in the complaints that Suno and Udio must have copied and “ingested” popular recordings in order to be able to generate material in particular styles.

“Building and operating a service like Suno’s requires at the outset copying and ingesting massive amounts of data to 'train' a software 'model' to generate outputs,” the labels allege in the complaint against that company. “For Suno specifically, this process involved copying decades worth of the world’s most popular sound recordings and then ingesting those copies into Suno’s AI model so it can generate outputs that imitate the qualities of genuine human sound recordings.”

The complaint against Udio makes similar allegations.

The labels specifically say in the complaint against Suno that the prompt “1950s rock and roll, rhythm & blues, 12 bar blues, rockabilly, energetic male vocalist, singer guitarist” together with the lyrics from the Chuck Berry song “Johnny B. Goode,” resulted in the service returning 29 outputs that “contain the style” of the original -- including one that “replicates the highly distinctive rhythm of the original’s chorus, and uses the same melodic shape on the phrases 'go Johnny, go, go.'”

“These similarities are only possible because Suno copied the copyrighted recordings that contain these musical elements,” the labels assert.

The lawsuits could face uphill battles in court, according to Santa Clara University law professor Eric Goldman.

He says one potential problem for the record labels is that the labels' allegations appear to be based on inferences about the companies' procedures, as opposed to direct proof -- such as admissions by company executives.

For instance, the Udio complaint contains the following passage: “It is obvious what Udio’s service is trained on ... Udio’s product can only work the way it does by copying vast quantities of sound recordings from artists across every genre, style, and era.”

The labels also say that before the lawsuits were filed, company executives failed to deny copying sound recordings.

“When Plaintiffs directly accused Udio of copying Plaintiffs’ sound recordings to train its model, Udio did not deny or proffer any facts to undermine those allegations,” the labels. “It would have been simple for Udio to say that it used other, legally acquired recordings, if that were the case. Instead, Udio deflected and disingenuously asserted that its training data is “competitively sensitive” and constitutes “trade secrets.”

The labels make similar allegations regrading Suno.

But neither complaint alleges that Udio or Suno executives ever admitted that they used copyrighted music to train the services.

Goldman says another potential obstacle for the record labels is that artificial intelligence companies may have a fair use defense to the allegations.

The record labels say in their complaints that fair use principles shouldn't apply, arguing that the concept “promotes human expression by permitting the unlicensed use of copyrighted works in certain, limited circumstances,” and that Suno and Udio are offering “imitative machine-generated music -- not human creativity or expression.”

Courts haven't yet definitively ruled on whether artificial intelligence companies have a fair use right to use copyrighted materials for training purposes.

But in 2015 a federal appellate court cleared Google of infringing copyright by digitizing books and displaying excerpts of them in its search engine. The 2nd Circuit Court of Appeals ruled in that case that Google's book project was "transformative," and therefore protected by fair use principles.