Logo via Spotify

Spotify has announced a new position on AI music, encouraging artistic use by cracking down on “slop.” Initially announced through a press conference with executives Charlie Hellman and Sam Duboff on Tuesday and elaborated with a blog post, Spotify’s new platform outlines three new approaches to regulating AI-generated uploads: A strengthened impersonation policy, a new spam filter and AI disclosures in song credits.

“We envision a future where artists and producers are in control of how or if they incorporate AI into their creative processes,” Spotify detailed in today’s post “As always, we leave those creative decisions to artists themselves while continuing our work to protect them against spam, impersonation, and deception, and providing listeners with greater transparency about the music they hear.”

Spotify’s revised impersonation policy builds on previous efforts to prevent “content mismatch” by implementing a new system that opposes unauthorized vocal deepfakes. While affording an exception to cases voices legally licensed to generative AI projects, the streaming platform took aim at a new wave of exploitative voice clones, establishing clearer recourse for impersonated artists. The platform has also announced new measures to prevent fraudulent uploads to popular artists’ profiles by working with major distributors. 

The spam filter, arriving over the coming months, is intended to curb the deluge of “mass uploads, duplicates, SEO hacks, artificially short track abuse, and other forms of slop [that] have become easier to exploit as AI tools make it simpler for anyone to generate large volumes of music.” The filter will identify and tag manipulative uploads to remove them from Spotify’s recommendation algorithm and “prevent spammers from generating royalties,” though they won’t be removed from the platform, and the tool will be implemented “conservatively” through the fall. According to today’s release, the company has removed 75 million spammy tracks in the last 12 months.

Finally, to properly identify uploads created with AI, Spotify will implement a new standardized AI disclosure with the music metadata provider Digital Data Exchange (DDEX). By the new standard, artists will be encouraged–though not required–to indicate what role AI played in a song’s creation, from AI-generated vocals to instrumentation, composing, post-production or lyrical editing.

“We know the use of AI tools is increasingly a spectrum, not a binary, where artists and producers may choose to use AI to help with some parts of their productions and not others,” the company detailed. “The industry needs a nuanced approach to AI transparency, not to be forced to classify every song as either ‘is AI’ or ‘not AI.’”

Spotify’s new AI policy arrives as the streaming giant faces scrutiny for CEO Daniel Ek’s venture capital firm Prima Materia and its investment in Helsing, a defense technology company specializing in AI-enhanced weapons and surveillance systems. Widespread objection to the connection has led several high-profile artists to remove their music from the platform, including Massive Attack, King Gizzard & The Lizard Wizard, Deerhoof, Xiu Xiu, Godspeed You! Black Emperor, Cindy Lee, Hotline TNT, Skee Mask and more.

“The economic burden that has long been placed on artists is now compounded by a moral & ethical burden,” Massive Attack wrote last week, “whereby the hard-earned money of fans & the creative endeavours of musicians ultimately funds lethal, dystopian technologies.”