Source : @Variety
Spotify has removed 50 million AI-generated “slop” tracks from listeners’ recommendations this year as the streaming giant steps up its campaign against low-quality and spam content created with artificial intelligence, Variety reported on Tuesday (Oct 20).
The figure was disclosed in a post on the company’s “For the Record” blog, in which Spotify set out the measures it has taken since it pledged to push back against the various forms of AI slop and spam that have spread across the platform. The 50 million tracks remain on Spotify but are no longer recommended to users. They come on top of more than 75 million tracks that were removed from the service entirely after being flagged as spam.
“Earlier this year, we looked closely at the tracks listeners were skipping, ignoring, or telling us they didn’t want to hear,” the company said in the post. “A clear pattern emerged: these tracks were consistently ones identified as slop or spam. Over the past six months, we’ve made numerous updates to our recommendation system to remove slop and spam, and surface more music made by authentic, human artists.”
Spotify added that its editors hold the same standard and do not programme content identified as slop or spam. “Since we made these changes, we’ve seen a marked decrease in listener concerns about this content showing up in recommendations,” it said.
The update also reported progress on several other measures aimed at AI content. Listeners can now see in a track’s credits which elements of a song were produced with AI, and Spotify said “thousands” of profiles now carry its “AI Persona” label. Those AI characters are also excluded from recommendations unless a user follows their profile. In addition, thousands of artists have opted into a programme that lets them screen which songs using their name appear on their profiles, an attempt to block the surge of AI-modified versions of songs that trade on the original musicians’ names.
The company also revealed that AI-generated songs are ten times as likely as other tracks to fall short of the 1,000-stream threshold required to earn royalties on the platform.
“This work isn’t finished,” Spotify said. “AI moves fast, and bad actors adapt just as fast. But as AI makes music more abundant, we believe listeners will value human artistry even more. Our job is to make Spotify the place they find it.”
The Swedish company first set out its AI protections in September 2025, when it announced a stricter impersonation policy covering unauthorised AI voice clones, a new music spam filter to identify and stop recommending mass uploads, duplicates and artificially short tracks, and support for an industry-standard DDEX credit system that lets artists and labels disclose how AI was used in a recording. At the time, Spotify said it had removed more than 75 million spammy tracks in the preceding 12 months, and noted that total music payouts on the service had grown from US$1 billion in 2014 to US$10 billion in 2024, making it a target for bad actors seeking to divert royalties.
The latest figures come as several streaming services move to make it clearer which songs in their libraries are produced by AI. According to Variety, YouTube Music last week joined Apple Music and Tidal in introducing labels for AI-powered tracks, while Tidal has also pledged not to allow monetisation of any AI-enabled song on its platform.
The original Variety post can be viewed on X: https://x.com/Variety/status/2112514190585782705.
About The Author
