Nearly Half Of All Songs Uploaded To Music Streaming Platform Deezer Are Now AI-Generated

Deezer said AI-generated uploads have surged to 75,000 per day, or 44% of all daily uploads, up from 10,000 in January 2025. AI music still represents only 1%-3% of total streams, but 85% of those streams are flagged as fraudulent and demonetised, underscoring monetization and integrity risks for the streaming ecosystem. The article also highlights a separate AI music fraud case involving over $8.1 million in alleged royalty theft.

Analysis

The economically important point is not that AI is flooding the catalog; it is that distribution rails are now becoming the bottleneck. When upload volume scales far faster than human review and takedown processes, platforms with better detection, moderation, and rights-clearance tooling can lower legal risk while simultaneously monetizing scarce legitimate inventory. That creates a subtle winner-take-more dynamic for incumbents with data, machine-learning, and label relationships, while smaller platforms face a rising fixed-cost burden just to keep their catalogs credible.

The second-order effect is on pricing power in the long tail of music rights. If AI-generated content continues to soak up upload capacity but only a small fraction of listening monetizes, the pressure shifts toward tighter royalty rules, fingerprinting standards, and fraud enforcement. That is structurally negative for low-quality content farms and any intermediaries whose economics depend on volume rather than verified engagement; it is more neutral-to-positive for major labels, publisher-admin platforms, and anti-fraud vendors because the industry will pay up for verification and enforcement.

The near-term catalyst set is regulatory and legal rather than consumer demand. The market is still underestimating how quickly rights holders will push for provenance labeling, watermarking, and platform liability, especially after a visible fraud case establishes the template for enforcement. Over 6-18 months, the risk is that streaming platforms adopt stricter upload gating and monetization rules, which could reduce top-line growth in total tracks but improve revenue quality and lower royalty leakage.

The contrarian view is that the headline growth in AI uploads may be less bearish for the music ecosystem than it appears, because most of the value accrues only when AI music is paired with fraud. If fraud detection keeps improving, AI becomes more of a moderation expense than a revenue threat. The bigger mispricing may be in vendors exposed to identity verification, content authenticity, and automated rights management, where this problem is just beginning to drive budget allocation.

Key Decisions for Investors

• Long SPOT vs short a basket of lower-quality digital media / UGC platforms over 3-6 months: if Spotify can better police AI/fraud and preserve advertiser confidence, it should outperform peers that face higher moderation costs and weaker trust economics.
• Accumulate RBLX or UGC-platform shorts only on rallies tied to 'AI content' enthusiasm; use 1-3 month horizons because any regulatory tightening around provenance can hit gross engagement assumptions quickly.
• Long RELY / PYPL-style fraud and identity verification exposure via a basket or options over 6-12 months: the market is underpricing budget migration toward fraud detection, provenance, and payment controls as platforms respond to AI-generated abuse.
• Pair long major labels (SONY / WMG) against short smaller streaming-adjacent monetization names over 6-12 months: tighter enforcement should favor rights owners with leverage while compressing economics for low-quality content distributors.
• For event-driven positioning, buy medium-dated calls on platform compliance winners after the next high-profile AI-fraud enforcement headline; risk/reward is favorable because adoption of stricter tooling can re-rate these names without needing consumer demand to improve.