Spotify Removed 50 Million AI Slop Tracks From Recommendations

Spotify has removed 50 million AI-generated tracks from its recommendation systems this year, the company disclosed Tuesday in a post on its "For the Record" blog. The figure sits on top of more than 75 million tracks the platform pulled entirely after flagging them as spam.

The 50 million tracks haven't been deleted. They remain available on the platform, but only for listeners who actively search for them. Spotify's algorithms and its human editors no longer push them into playlists or personalized feeds.

"Earlier this year, we looked closely at the tracks listeners were skipping, ignoring, or telling us they didn't want to hear," Spotify wrote in the post. "A clear pattern emerged: these tracks were consistently ones identified as slop or spam."

The company said it has updated its recommendation system repeatedly over the past six months to filter out that content and, in its words, "surface more music made by authentic, human artists." Its editors now apply the same standard and don't program anything identified as slop or spam. Since the changes, Spotify reports "a marked decrease in listener concerns about this content showing up in recommendations."

What else has Spotify built to contain AI music?

The blog post recapped a set of tools the company has rolled out alongside the recommendation purge:

• AI credits disclosure. Listeners can now see which elements of a song were produced with AI within a track's credits.
• AI Persona labels. Spotify says "thousands" of profiles are now tagged as AI Personas — virtual characters whose music is excluded from recommendations unless a user follows the profile directly.
• Artist name screening. Thousands of artists have opted into a program that lets them control which songs bearing their name appear on their profiles, a defense against AI-modified variants of songs that trade on original musicians' names.

One more number stands out. AI-generated songs are ten times as likely not to reach the 1,000-stream threshold required to earn royalties, Spotify revealed. In other words, machine-made tracks overwhelmingly fail the platform's lowest commercial bar.

What does this change for artists and listeners?

For listeners, the practical shift is already here: less AI filler in Discover Weekly-style feeds and editorial playlists, with the slop still technically searchable. For working musicians, the artist screening program offers a first line of defense against AI clones cashing in on their names — and the royalty data suggests the economics of mass-produced AI music remain thin.

Spotify itself frames this as a permanent maintenance job rather than a solved problem. "This work isn't finished," the company wrote. "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 wider industry is moving the same way

Spotify's disclosure lands amid a broader push for transparency across streaming. YouTube Music last week joined Apple Music and Tidal in introducing labels for AI-powered tracks. Tidal has gone a step further and promised not to allow monetization for any AI-enabled song on its platform.

With three major services now labeling AI content and Spotify quantifying its purge in the tens of millions, disclosure of machine-made music is shifting from experiment to industry baseline — and the next battleground looks like monetization rules, where Tidal currently holds the hardest line.