Can Spotify Regulate Misleading AI-Generated Music?

"For artists unaffiliated with a record label or publishing house, online services can provide a leg up and a means to get one's work in front of a potentially vast audience. That's the good news; the bad news is that it's also made it easier for unethical people to use the same services to flood potential listeners or readers with, for lack of a better phrase, slop made using AI.This has become a growing concern in the book world, and it's also present in the music industry."

"Futurism's Victor Tangermann recently reported that the Spotify page for Volcano Choir, a project featuring Bon Iver's Justin Vernon, features a "new" song that was almost certainly created by AI and not any of the actual members of Volcano Choir. As the service itself recently acknowledged, this was symptomatic of a much larger issue.This week, Spotify announced a shift in its policy when it comes to generative AI and music. "[W]e believe that aggressively protecting against the worst parts of Gen AI is essential to enabling its potential for artists and producers," the company said in its statement. It also noted that it has removed 75 million "spammy tracks" from the service in the last year."

"Among the steps Spotify is implementing are stronger restrictions addressing voice impersonation, including "testing new prevention tactics with leading artist distributors." Spotify also plans to support AI disclosure and "a new music spam filter" that they believe will help identify bad actors who repeatedly upload illicit AI-generated music."

Online platforms give independent artists greater access to audiences while also enabling unethical actors to flood services with AI-generated low-quality content. An apparent AI-created song appeared on a well-known project's streaming page, reflecting a broader problem across music and publishing. Spotify acknowledged the issue, reported removal of 75 million spammy tracks, and announced policy changes targeting generative AI in music. Measures include stronger restrictions on voice impersonation, testing prevention tactics with distributors, supporting AI disclosure, and deploying a music spam filter to identify repeat illicit uploaders.

Read at InsideHook