Spotify has removed more than 50 million songs identified as low-quality content, spam or mass-produced material created with artificial intelligence from its recommendation systems. The move is part of the streaming platform’s efforts to improve music discovery and give greater visibility to artists whose work may be overshadowed by the growing volume of automatically generated tracks.
However, it is important to distinguish between removing songs from recommendations and deleting them from the platform. Most of the tracks affected by this measure have not been permanently removed from Spotify’s catalog. Instead, they have been excluded from algorithmic recommendations and other music discovery features, meaning users may still be able to find and play them through direct searches.
Separately, Spotify reported that it had completely removed more than 75 million songs for serious or repeated violations of its anti-spam policies. The figures refer to two different measures: restricting the visibility of certain tracks and removing content that violates the platform’s rules.
Why Spotify is taking action against AI-generated music
The decision followed an analysis of users’ listening habits. According to Spotify, songs that listeners frequently skip, ignore or flag as unwanted often overlap with material previously identified as spam or low-quality content.
Over the past six months, the company has adjusted its recommendation systems to reduce the circulation of these tracks and prioritize music by verified human artists. It has also instructed its editorial teams not to include material classified as spam or low-quality content in playlists curated by the platform.
The growing availability of generative AI tools has made it easier to produce music in large quantities and upload tracks to streaming services. Although AI can offer new creative possibilities, mass distribution can also be used to flood platforms with tracks designed to generate streams rather than provide meaningful musical experiences.
This practice is often associated with the term slop, which describes mass-produced AI-generated content perceived as having little value. The phenomenon has raised concerns about the quality of recommendations and the visibility of musicians who invest time and resources in creating and promoting their work.
The platform’s strategy focuses on identifying abusive practices and low-quality material rather than banning all music created with artificial intelligence.
How Spotify identifies AI-generated songs and spam
Spotify uses multiple signals to detect suspicious content rather than relying on a single method. Its systems analyze listener behavior, upload frequency and patterns, recording metadata, and information provided by external AI-detection tools.
The platform also monitors tactics used to manipulate its systems, including mass uploads, duplicate recordings, search result manipulation, fraudulent playlist networks and unusually short tracks. These practices can be used to generate artificial engagement or exploit recommendation and royalty distribution systems.
The company acknowledges that identifying AI-generated music remains a challenge. As generative technology advances, distinguishing between a recording created by a person and one produced by a machine becomes increasingly difficult. Detection tools also have limitations, making it necessary to combine different indicators and continuously update monitoring systems.
This distinction matters because not every AI-generated song constitutes spam, and using this technology does not automatically mean a track violates Spotify’s rules. The platform’s stated goal is to combat content and distribution practices that harm the listening experience or breach its policies.
New tools to improve transparency and protect artists
Alongside changes to its recommendation systems, Spotify has introduced tools designed to provide greater transparency about the role of artificial intelligence in music production.
One of these initiatives is AI Credits, which allows artists to disclose which parts of a song were created with the assistance of artificial intelligence. The information is intended to help listeners better understand how a recording was produced. According to Spotify, the platform receives tens of thousands of these disclosures each day.
The company has also introduced AI Persona, a label for music profiles whose artistic identities were created entirely with AI. By default, these profiles are excluded from editorial and algorithmic recommendations, although users may still discover their music by searching for or following them.
Another initiative is Verified by Spotify, a verification system designed to help listeners identify profiles associated with real, identifiable people. The company has reported that more than 300,000 artists have received this verification, including numerous independent musicians.
In addition, Spotify is testing a tool that allows certain artists to review and block releases that use their names without authorization before those tracks appear on their profiles. The initiative aims to address impersonation and other forms of misuse that can damage an artist’s reputation or divert attention from legitimate releases.