AI-generated tracks now account for 34% of daily uploads to streaming platforms, diluting the royalty pool for human artists

Deezer reported receiving approximately 50,000 fully AI-generated tracks per day by late 2025, accounting for 34% of all daily uploads. Across all platforms, an average of 106,000 new tracks were delivered to streaming services each day in 2025. While Spotify removed over 75 million 'spammy tracks' and reports that AI content accounts for only 0.5% of actual streams, the sheer volume of AI uploads creates a structural problem: the royalty pool is a fixed percentage of platform revenue, and every stream captured by AI-generated content is a stream — and a fraction of a penny — that does not go to a human artist. The damage is not primarily about AI tracks going viral and stealing listeners. It is about the long tail. AI generators can produce thousands of tracks per day targeting high-value functional niches: lo-fi study beats, ambient work playlists, meditation soundscapes, coffee shop jazz. These tracks individually capture modest stream counts but collectively siphon millions of streams from human artists who create in those same genres. A human lo-fi producer who spends 40 hours on a track competes against an AI that produces 500 similar tracks in an afternoon, each one snagging a few hundred streams from algorithmic playlist placement. The human artist does not lose in a dramatic, visible way — they lose $0.002 at a time, across thousands of micro-competitive moments they cannot even see. For artists in ambient, classical, jazz, and instrumental genres, this slow erosion can represent 10-20% of their streaming income without them ever understanding why their numbers declined. This problem persists because platforms have conflicting incentives. More content means more listening hours, which means more ad revenue and subscriber retention. Platforms benefit from a flooded catalog even if individual creators suffer. Detection is also genuinely hard: the line between 'AI-assisted production' (which most modern music involves to some degree) and 'fully AI-generated content' is blurry and getting blurrier. No platform has implemented mandatory AI disclosure labeling, and the legal framework for AI-generated content's copyright status remains unsettled. Distributors like DistroKid and TuneCore have financial incentives to accept as many uploads as possible, since they charge per-track or per-year fees. The entire pipeline — from generation to distribution to streaming — rewards volume over artistry.