Spotify deletes millions of spam songs and adds AI labels
· re:union Radio editorial · 1 min read
Spotify tackles spam uploads by deleting or filtering millions of tracks, while introducing new voluntary transparency tools for AI-generated music.
Spotify is intensifying its measures against spam and mass uploads. In the twelve months prior to September 2025, the streaming platform removed over 75 million tracks classified as spam. Furthermore, in 2026, the service excluded more than 50 million additional tracks from its recommendation algorithms and editorial playlists.
Protecting royalties and adding transparency
Roughly 100,000 new tracks arrive on the service's servers every single day. Because mass uploads directly affect payout distributions, these removals are intended to protect the royalty pool. In 2025, the company paid out more than 11 billion US dollars in royalties.
The service sorts out unwanted content using listening habits, upload patterns, metadata, and external AI detection tools. According to platform data, pure AI music accounts for significantly less than one percent of total consumption.
Several transparency features were launched in April 2026:
- Through the beta feature AI Credits, artists can voluntarily declare whether AI was utilized for vocals or instrumentation.
- Thousands of AI-generated identities have received an AI Persona Badge, excluding them from algorithmic and editorial recommendations unless users follow them directly.
- The badge Verified by Spotify was introduced to confirm the authenticity of artist profiles.
Industry bodies including IFPI and RIAA have also presented a voluntary labeling system distinguishing between AI-generated and AI-assisted works.
Source: Delamar
