The Dark Side of AI Music: How Bots Are Stealing Millions From Real Artists — Deezer’s Alarming New Report

Key Takeaways:

• 70% of AI music streams on Deezer are fraudulent, despite AI tracks representing only 0.5% of total platform streams
• Fraudsters use AI to generate thousands of tracks and deploy bot networks to artificially inflate listening numbers for royalty payments
• The scale of fraud has reached industrial levels, with one case involving over $12 million in fraudulent earnings and 661,000 daily fake streams

Deezer has revealed a startling reality about AI-generated music fraud: 70% of streams for AI-created tracks on its platform are fraudulent, representing a sophisticated scheme where artificial intelligence creates both the content and fake listeners. While AI music comprises just 0.5% of total streams, this manipulation targets royalty payments through automated listening farms.

The French streaming giant processes approximately 20,000 AI-generated tracks daily—18% of all uploads—using detection tools that identify content from generators like Suno and Udio with 100% accuracy. Thibault Roucou, Deezer’s director of royalties and reporting, emphasized the financial motivation: “As long as there is money [in fraudulent streaming] there will be efforts, unfortunately, to try to get a profit from it”.

This revelation follows the 2024 arrest of Michael Smith, who allegedly defrauded streaming services of $12 million using 1,040 bot accounts generating over 661,000 daily streams. His operation demonstrates the industrial scale possible: earning an estimated $1.2 million annually at half-cent-per-stream royalty rates.

The crisis extends beyond simple copyright infringement, creating an ecosystem where AI simultaneously produces and consumes content, fundamentally distorting streaming economics and artist compensation structures.

This industrial-scale fraud directly impacts legitimate musicians, as Grammy-nominated artists have discovered when finding fake AI tracks attributed to their names on streaming platforms.