Spotify declares war on AI slop, banning over 50 million tracks from recommendations

As AI-generated content continues to flood digital platforms, Spotify, the world's largest music streaming platform, has declared war on AI slop — low-quality content produced by artificial intelligence — announcing it will remove more than 50 million such tracks from its recommendation lists.

Spotify Korea is expected to follow the global headquarters' policy, signaling that the pushback against AI slop is gaining momentum worldwide.

Spotify announced Wednesday on its official blog that it would remove more than 50 million AI-generated tracks from its user recommendation lists.

The move means AI slop will be excluded from the algorithms that suggest content to listeners. The decision carries significant weight for Spotify, a platform where personalized features — such as autoplay and home screen recommendations — are central to the user experience.

Spotify described the tracks as "garbage and cheap," saying they have "no reason to remain on our lists any longer."

Regulatory pressure against AI slop is also intensifying globally. French music streaming platform Deezer said Tuesday it had detected and removed more than 42 million AI-generated tracks from its song recommendation lists.

Deezer also plans to introduce a toggle feature by year's end that will let users choose whether AI-generated tracks appear in their search results.

Spotify ranked second in South Korea's monthly active user count for music streaming apps in June, with about 6.22 million users, trailing only YouTube's roughly 9.49 million, according to data from Wise App & Retail. As the platform tightens its stance on AI slop, the industry is watching closely to see whether the move will ripple through the domestic music streaming market.

Spotify has already been taking steps to address AI content, including labeling such material on an ongoing basis. It has also restricted royalty payments for tracks with abnormally high streaming revenue by screening them for AI generation.

User sentiment is largely negative. A survey on AI labeling guidelines conducted by research tech firm OpenSurvey, which polled 3,000 men and women nationwide and 900 SNS users, found that 85.6 percent of respondents said AI-generated content should be labeled, with 54.1 percent saying such labeling is absolutely necessary.

Public frustration has also been fueled by a recent revelation that "Kkumkkudeon Eoreuni Doeeonnayo" — a song meaning "Did you become the adult you dreamed of?" — had won wide affection among listeners before it emerged the track was AI-generated. Comments on music platforms ranged from "I had no idea" to "I'm genuinely shocked" and "Isn't this disgusting?"

However, current classification and labeling standards for AI-generated music apply only to businesses, not individual creators, leaving platforms such as YouTube and Melon reluctant to take preemptive action.

"Discussions about AI-generated works have been so heated in the industry lately that companies have no choice but to tread carefully," an industry official said. "Standards need to be established for how to classify and label AI-generated music."