AI-Generated Music Floods Platforms: Deezer Sees Over Half of Daily Uploads Synthetic

A surge in AI-generated music uploads - Deezer's June figure of ~90,000 daily synthetic tracks - is changing the economics and curation dynamics of streaming platforms. On one hand, synthetic content democratizes creation, scales catalog size, and enables niche personalization. On the other, it floods distribution channels with variable-quality material, complicates rights management, and heightens the risk of undisclosed use of copyrighted material in training datasets.

For labels and publishers, the risks are clear: potential revenue dilution, harder attribution, and increased monitoring costs to detect infringement. Platforms must decide whether to welcome synthetic tracks as user-generated content or to gate them, both choices carrying trade-offs in user engagement and legal exposure. Recommendation systems also face dilution: more tracks mean sparser signals per item, increasing the importance of robust metadata, provenance flags, and quality signals to preserve user experience.

Practical steps for leaders include implementing mandatory machine-readable provenance metadata for uploads, adopting or developing AI-origin detection tools, and updating licensing terms to cover synthetic works explicitly. Platforms should also pilot differentiated UX and monetization models - for example, separate discovery lanes for verified artist content and synthetic creations, dynamic payout rules, and subscription features for AI-driven personalization.

Longer-term, industry coordination on standards for dataset licensing, watermarking, and provenance will be essential. Companies that proactively adopt technical controls and commercial rules around AI-generated content can protect IP, maintain trust with creators, and unlock new product opportunities (custom tracks, on-demand composition) while mitigating the downside of an increasingly synthetic catalog.