The music streaming world is facing an unprecedented deluge of artificial intelligence-generated tracks, with platforms reporting tens of thousands of such songs uploaded daily. According to data from Deezer, a Paris-based streaming service, AI-created music now accounts for 44 percent of all downloads on its platform, a sharp rise from earlier figures that highlights the rapid encroachment of automated content into an industry long dominated by human creativity.
Deezer's in-house AI-detection software has been instrumental in tracking this surge. In January 2025, the company identified 10,000 AI-generated tracks uploaded each day. By April of that year, the number had climbed to 15,000, reaching 30,000 in September and 50,000 by November. As 2026 began, uploads hit 60,000 per day in January, and the most recent figures show 75,000 daily—a number that experts say is likely already outdated given the pace of technological advancement. In contrast, human-created songs remain steady at around 95,500 per day, a figure that has shown little change over the past 15 months.
This phenomenon isn't isolated to Deezer. Industry observers believe all major platforms, including Spotify, Apple Music, and YouTube Music, are grappling with similar volumes. Deezer has proactively offered its detection software to other streamers, having already identified 13.4 million AI tracks since the start of 2025. The ease of creation is a key driver: tools like Suno, Google Magenta, Loudly, and Mubert allow users to generate full songs from simple text prompts, delivering results in seconds.
Beyond the sheer volume, the influx raises significant concerns about energy consumption. Each AI-generated track requires between 0.001 and 0.01 watt-hours of electricity for creation alone. While that may seem negligible for a single song, multiplied across 75,000 daily uploads, the cumulative energy use becomes substantial. This doesn't even account for the power needed to store and transmit these files on servers or the additional drain from automated bots that simulate streams to inflate popularity and royalties.
Streaming fraud has emerged as a particularly insidious byproduct. Last month, Michael Smith, a 54-year-old resident of North Carolina, pleaded guilty to orchestrating a multi-year scheme that exploited AI music generation. Over seven years, Smith used AI to produce hundreds of thousands of low-quality tracks, uploading them to various platforms. He then deployed networks of bots across thousands of automated accounts to artificially stream the songs, making the activity hard to detect.
At its peak, Smith's operation generated over 660,000 streams per day, translating to approximately $1.2 million in annual royalties siphoned from the shared payout pool that funds legitimate artists. The scheme deprived human musicians of earnings that rightfully belonged to them under standard royalty distribution rules. Smith was apprehended in September 2024, and as part of his plea, he forfeited more than $8 million in ill-gotten gains. He faces a maximum penalty of five years in prison, with sentencing scheduled for July 29.
Human artists are suffering direct harms as well. Paula Toledo, a singer based in Vancouver, Canada, discovered that her original music had been co-opted and re-released under fraudulent accounts controlled by unknown perpetrators. "I had all kinds of trouble proving to Spotify, Apple, and others that I was the rightful owner of that music," Toledo said, highlighting the bureaucratic hurdles in reclaiming her work.
Similarly, Grace Mitchell, an American-Australian singer-songwriter, was stunned to find her songs stolen and re-monetized by unidentified fraudsters. "It was shocking to discover that my music had been hijacked," Mitchell recounted, emphasizing how royalties intended for her were diverted elsewhere. These cases underscore a growing pattern where AI tools not only flood the market with new content but also facilitate the unauthorized repurposing of existing human creations.
Even established artists aren't immune. Reports have surfaced of legitimate albums, such as those by Van Morrison, being re-uploaded under pseudonyms to game the system. A singer-songwriter friend of a Global News reporter recently faced the same issue, discovering her catalog hijacked without permission. Such incidents erode trust in streaming platforms and complicate the verification process for creators.
Fake artists, entirely fabricated by AI, are also topping charts and misleading listeners. IngaRose, an AI-generated persona, reached the top of the iTunes sales charts in Canada, the United States, the United Kingdom, France, and New Zealand earlier this year with the track "Celebrate Me." The song appeared in nearly 300,000 TikTok videos, amplifying its reach. Likewise, Breaking Rust, another AI creation, claimed the number-one spot on the Billboard Country Digital Song Sales chart in November with "Walk My Walk."
Other examples include Xania Money, SiennaRose, and Eddie Dalton, all nonexistent performers whose discographies are purely digital inventions. Last year, The Velvet Sundown, a simulated California folk rock band, fooled audiences with its generic sound. More recently, Promptgenix released three albums in just two months, lacking any biographical details. Investigations revealed it was generated by an AI engine called Promptgenie, created by an individual frustrated with human music offerings.
A survey from November—conducted amid the accelerating evolution of AI music—found that 97 percent of listeners couldn't distinguish AI-generated tracks from human ones. Despite widespread disapproval of software-created art, a 2025 Luminate report indicated that up to 50 percent of U.S. listeners are unbothered by it. This ambivalence could further entrench AI's role in the industry.
Major labels are responding with a mix of litigation and collaboration. Warner Music Group settled lawsuits against Suno and Udio, AI music platforms, and entered into an undisclosed partnership. Universal Music Group, after initial attempts to sue Udio out of existence, also struck a deal with the company. The long-term effects of these arrangements remain unclear, but they signal a potential shift toward integrating AI, which could reshape opportunities for human musicians.
The parallels to earlier eras are striking. Decades ago, Memorex's advertising campaign posed the question, "Is it live or is it Memorex?" Today, as AI blurs the lines between human and machine creativity, a new dilemma confronts the industry: Is it living or is it software? With uploads accelerating and detection tools playing catch-up, the music world braces for deeper transformations.
Looking ahead, the proliferation of AI music could exacerbate inequalities, diverting resources from genuine artists while environmental costs mount. Streamers like Deezer continue to advocate for shared solutions, but without broader regulation, the "tsunami" of automated content shows no signs of receding. As one industry watcher put it, the situation is only going to get weirder, challenging the very essence of musical authenticity in the digital age.