Deezer disclosed on July 21, 2026 that AI-generated music now accounts for more than half of all daily uploads to its platform. The figure marks a rapid escalation from 10% just eighteen months earlier, when the streaming service first began tracking the metric in January 2025 (TechCrunch).

The trajectory is steep and consistent. Deezer reported 10,000 AI-generated tracks per day in January 2025 (10% of uploads), rising to 20,000 (18%) by April, 30,000 (28%) by September, and 50,000 (34%) by November. By January 2026 the figure reached 60,000 (39%), and by April 2026 it climbed to 75,000 (44%). Daily AI uploads peaked in June 2026 at a monthly average of 90,000 tracks per day, according to the company's data. Across all of 2025, Deezer detected and tagged 13.4 million AI-generated songs (Deezer).

Deezer CEO Alexis Lanternier stated that "half of all daily uploads are AI-generated tracks" and that the company is taking additional steps to safeguard the rights of artists and songwriters. Those steps include a takedown policy: Deezer will remove AI-generated tracks that have not been streamed in the preceding six months or that are implicated in fraudulent streaming activity. AI-generated songs are also systematically excluded from Deezer's Flow recommendation feature and its algorithmic playlists (Deezer).

The detection technology underpinning these measures can identify tracks produced with models from Suno and Udio, the two most prominent AI-music generation startups. Deezer claims 99.8% accuracy for its detector, which it made available to rival streaming platforms in January 2026. In June 2026, the company released a free tool that can identify AI-generated tracks in playlists from Spotify, Apple Music, and other services, launching it publicly on June 11 (Reuters). Deezer asserts it is the only streaming platform actively detecting and tagging AI-generated music at scale (Deezer).

The competitive landscape has fragmented. Bandcamp banned AI-generated music outright in January 2026. Tidal cut off monetization for AI-generated tracks in June 2026. Deezer has taken a middle path: labeling rather than banning, but actively filtering AI content out of recommendation surfaces and removing low-engagement or fraudulent tracks. The company began labeling AI-generated music on its platform in June 2025 (TechCrunch).

Deezer had 9.7 million subscribers as of November 2025 (Reuters). The platform is not among the largest streaming services by subscriber count, but its detection infrastructure and public reporting on AI upload volumes have made it a primary data source for an industry-wide problem that Spotify and Apple Music have not addressed with comparable transparency.

The growth curve itself deserves attention. The share of AI-generated uploads roughly doubled every six months from early 2025 through mid-2026. If that rate of growth has plateaued at the June 2026 peak of 90,000 daily uploads, it may signal that AI-music generation tools have reached a saturation point among their user base, or that platform-level countermeasures are beginning to suppress the inflow. The data does not yet distinguish between these explanations.

Worth flagging is what the industry is not measuring. Deezer can detect output from Suno and Udio models specifically. The detection landscape for other, newer, or fine-tuned models is not publicly documented. As the Reuters headline from November 2025 noted, a survey at the time found AI music to be "virtually undetectable" by human listeners. The arms race between generation and detection is structurally familiar to anyone who has watched adversarial dynamics in spam filtering, deepfake detection, or cybersecurity more broadly. Each side iterates; each iteration narrows the other's window of advantage.

The economic pressure is real. Streaming royalties are divided among rights holders based on play counts. When 50% of uploaded content is machine-generated, the denominator grows without a corresponding increase in listener demand, diluting per-stream payouts for human-created work. Deezer's decision to exclude AI tracks from algorithmic playlists and remove unstreamed AI content directly addresses this dilution mechanism, but only on its own platform. Spotify and Apple Music, which command far larger subscriber bases, have not disclosed comparable detection or filtering measures.

Looking at what this means for the broader ecosystem: the platforms that move first on detection and labeling may set de facto standards that larger competitors find difficult to ignore, particularly if rights-holder organizations, regulators, or litigation begin demanding evidence of good-faith effort to separate human and machine output. Deezer's decision to open-source its detection capability to rivals in January and release a free public tool in June suggests the company sees detection infrastructure as a strategic differentiator, not a proprietary moat.

The question for the next eighteen months is not whether AI-generated music will continue to flow onto streaming platforms. It will. The question is whether the platforms receiving it can build, maintain, and enforce detection systems that keep pace with generation tools that are improving on a similar cadence. Deezer's data shows the inflow. It does not, on its own, show that the outflow, filtering, and labeling are scaling at the same rate.