Overview

• Spotify says it removed more than 75 million spammy tracks over the past 12 months as AI tools accelerated mass, low-quality uploads.
• A new impersonation policy bans unauthorized AI voice clones and other vocal deepfakes unless the impersonated artist has authorized their use.
• The platform will roll out a music spam filter this fall that flags mass uploads, duplicates, SEO hacks and artificially short tracks, then stops recommending them, with a conservative deployment.
• Spotify is backing a DDEX-developed metadata standard for AI disclosures in credits and plans to display where AI was used as labels and distributors submit the information, with 15 partners committed.
• The company says engagement with AI-generated music is minimal and rejects claims that it adds or promotes AI tracks in playlists for financial benefit.