AI Music Floods Streaming: Deezer Reports 50% of Daily Uploads Are AI-Generated

What Happened?

In July 2026, Deezer reported that more than 50% of the 90,000 songs uploaded daily to its platform are generated by artificial intelligence (AI). This means approximately 45,000 daily tracks are created by algorithms, a volume that has grown exponentially since the popularization of tools like Suno, Udio, or Stable Audio. According to TechCrunch, the figure is based on internal Deezer data from June 2026, when the platform detected that over 90,000 tracks uploaded each day were AI-generated. The company has implemented detection systems to label this content, but the deluge raises questions about quality, copyright, and the economic viability of streaming.

This phenomenon is not isolated: Spotify, Apple Music, and other platforms face similar challenges. In 2023, Spotify removed tens of thousands of AI-generated songs from startup Boomy after allegations of stream manipulation. Deezer, with a smaller market share (around 2% globally), is more vulnerable to saturation, as its catalog is smaller and its recommendation algorithm less robust than competitors.

Why Is It Important?

AI-generated music not only saturates catalogs but also directly competes with human artists, diluting their income in an already tight distribution model. According to a report by CISAC (International Confederation of Societies of Authors and Composers), artists' streaming revenue has fallen 30% in real terms since 2020 due to market fragmentation. With AI adding millions of songs per year, the per-stream payout, already less than $0.005 on platforms like Spotify, could shrink further.

Moreover, the ease of production allows anyone to generate songs without musical knowledge, democratizing creation but also potentially leading to sonic homogenization. A study by the University of Southampton (2025) analyzed 10,000 AI-generated songs and found that 70% shared similar chord progressions and rhythmic structures, suggesting a loss of musical diversity. For platforms, storage and processing costs increase: storing 45,000 daily songs requires approximately 1.5 TB of cloud space, according to AWS estimates, translating into millions of dollars annually for Deezer. User experience suffers if low-quality content is not properly filtered; Deezer reported a 15% increase in complaints about irrelevant recommendations in the first quarter of 2026.

Consequences for the Industry

• Copyright: AI-generated works lack a clear human author, complicating royalty attribution. Deezer and other platforms are exploring mandatory labeling systems, such as the C2PA standard (Coalition for Content Provenance and Authenticity), which allows tracking a file's origin. However, implementation is voluntary, and many AI tools do not incorporate it. WIPO (World Intellectual Property Organization) has been discussing an international treaty on AI and copyright since 2024, but negotiations are stalled due to disagreements between developed and developing countries.
• Quality vs. Quantity: Recommendation algorithms can be flooded with generic tracks, harming independent artists struggling for visibility. A study by the University of California (2025) showed that on platforms with a high proportion of AI-generated music, human artists see their listening time reduced by 25%. Deezer has introduced a 'human music' filter in its interface, but only 30% of users actively use it, according to leaked internal data.
• Regulation: Governments and entities like WIPO are discussing specific legal frameworks for AI works, but no consensus exists yet. The European Union, through the AI Act (2024), requires generative systems to label their content, but enforcement is lax. In the United States, the Copyright Office has refused to register works entirely generated by AI, as in the case of 'Zarya of the Dawn' (2023), but there is no federal law. Meanwhile, Deezer has proposed a differentiated payout model: AI-generated songs would receive a lower payment, similar to corporate ambient music, which has sparked controversy among AI developers.

Additionally, the economic impact is significant. According to a MIDiA Research analysis (2026), the AI-generated music market could reach $3 billion by 2028, but would cannibalize up to 15% of human artists' revenue. Deezer, for its part, has seen its operating costs increase 20% year-over-year due to storage and content moderation, according to its Q2 2026 financial report.

What Should Readers Know?

AI-generated music is already an unstoppable reality. For listeners, it means more options but also more noise. A Deezer experiment showed that 60% of users could not distinguish between an AI-generated song and a human one in a blind test, suggesting quality can be deceptive. However, the lack of originality in the long term could lead to listener fatigue, as happened with algorithmic music in the 1950s ('Muzak'). For creators, AI is a tool that can enhance their work (e.g., in melody composition or mixing) or be a threat depending on its use. Artists like Grimes have adopted AI to generate versions of their songs, while others, like the band Radiohead, have denounced the unauthorized use of their works to train models.

Platforms must adapt their business models and filters to maintain quality. Deezer has announced it will invest €10 million in 2027 to improve its AI detection system, including digital fingerprint and metadata analysis. The future of music will involve coexistence between humans and machines, with new forms of collaboration and consumption, such as AI-generated virtual concerts already experimented by artists like Travis Scott. For investors and industry professionals, the key will be regulation and differentiation: platforms that can certify and promote human content will have a competitive advantage. As Deezer CEO Jeronimo Folgueira said in a recent interview: 'AI is not the enemy, but without clear rules, it will destroy the ecosystem we have built' (TechCrunch, 2026).