The company said it removed 50 million tracks, but hasn't necessarily wiped them entirely.

Just over a year ago, Spotify announced policy changes to address the tidal wave of AI-generated audio that has flooded its platform. In a blog post on Tuesday, the company provided an update on the initiative, but it's still not going far enough to help users avoid genAI trash.

So far in 2026, Spotify says it has yanked 50 million tracks it categorized as "slop and spam" from recommendations. Over the last six months, it made several updates to the recommendation system to instead promote more tracks from "authentic, human artists." The company's editors don't include tracks identified as slop or spam in curated playlists either.

These updates follow Spotify taking a close look at songs users "were skipping, ignoring or telling us they didn't want to hear." It determined there was a clear pattern of users rejecting tracks that were 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," Spotify said.

The company added it will have to keep updating its systems to stay on top of spammers, who will inevitably keep changing tactics to evade the rules as long as possible. Spotify said it uses many signals to help it detect slop and spam, such as metadata, user feedback, upload patterns and third-party AI detection tools. It noted that the latter tech has limitations and it'll become harder for those tools to detect slop as generative AI systems advance.

Source link