Spotify Purges 75 Million AI Tracks to Fight Royalty Fraud
Stockholm, Sunday, 26 July 2026.
Spotify has removed 75 million AI-generated spam tracks to stop bot-driven fraud, protecting royalty pools and ensuring legitimate artists receive their fair share.
The Mechanics of Royalty Exploitation and “AI Slop”
The massive purge of over 75 million tracks between July 2025 and July 2026 highlights a critical vulnerability in the music streaming ecosystem: the exploitation of the royalty pool model [1][2]. Under this model, Spotify distributes payout funds based on aggregate listening figures, making it highly susceptible to manipulation [2]. Fraudulent actors have increasingly utilized automated spam bots to upload tracks featuring fake metadata, generic titles, and artificially short durations [2]. This phenomenon, often described by industry experts as “AI slop,” refers to low-quality content produced in seconds and uploaded purely to manipulate the platform rather than serve listeners [2]. A primary example of this “catalog-stuffing” involves flooding the platform with thousands of 30-second tracks, such as “relaxing rain sounds,” designed to rapidly accumulate play counts and siphoning royalties away from human creators [1].
The Scale of Daily Uploads and Financial Impact
The scale of this issue is compounded by the sheer volume of daily uploads. Industry estimates indicate that approximately 100,000 new tracks are uploaded to Spotify daily, with nearly 50% estimated to involve some form of AI generation [2]. This means that roughly 50000 AI-influenced tracks are introduced to the platform every single day [2]. To combat the resulting deluge of spam, Spotify’s policy leadership has emphasized that spammers should not earn royalties that should rightfully go to professional artists and songwriters [1]. Consequently, as of late July 2026, the streaming giant continues to actively purge these low-quality, automated tracks to protect its royalty pools from bot-driven fraud [1][2].
Evolving Policies and Technological Countermeasures
To defend its ecosystem against these sophisticated manipulation tactics, Spotify has deployed a multi-layered technological framework. The platform utilizes a dedicated monitoring team to track evolving spam tactics in real time, alongside anti-scraping defenses designed to prevent generative AI companies from bulk-harvesting its catalog for model training [1]. Additionally, Spotify employs an advanced machine-learning filter specifically trained to identify and remove mass-uploaded or artificially streamed content [1]. This filter acts as a digital barrier against the automated distribution networks that generate and upload vast quantities of low-effort audio files [1][2].
Strict Policy Mandates and Historical Precedents
These technical defenses are reinforced by strict policy mandates. Spotify’s current policy requires the use of DDEX metadata standards to clearly label tracks as either “fully AI-generated” or “AI-assisted” [1]. The guidelines also strictly ban unauthorized voice cloning, safeguarding artists’ unique vocal identities from unlicensed replication [1]. This is not the platform’s first intervention; in 2023, Spotify removed tens of thousands of AI songs associated with the platform Boomy due to detected artificial streaming [1]. The platform subsequently updated its AI protection measures on September 25, 2025, to further target unauthorized voice clones and mass uploads [1][2].
The Delineation Between Utility and Spam
Crucially, Spotify’s aggressive enforcement does not represent a blanket ban on artificial intelligence in music creation. Instead, the platform actively distinguishes between malicious spam and legitimate creative tools [2]. Spotify’s policy explicitly permits authorized AI remixes, covers, and AI-assisted mixing or mastering [1]. Furthermore, the company continues to collaborate with major record labels to verify and clear legitimate AI-created content for commercial distribution [1]. The objective remains focused on preventing automated “slop” from gaming the payout system, rather than stifling technological innovation among human artists [1][2].
Industry-Wide Transparency and Labeling Initiatives
This platform-specific crackdown occurs amid a broader, industry-wide debate regarding AI transparency and intellectual property. The recorded music industry is currently advancing a proposal to mandate the labeling of AI-generated content on all streaming platforms [3]. This proposal emphasizes the need to distinguish traditional digital audio workstation (DAW) tools—such as pitch correction, MIDI drums, and virtual studio technology (VST) orchestral plugins—from fully automated, human-less prompt-based generation [3]. Concerns have particularly focused on platforms like Suno, where free membership tiers allow users to generate and upload complete songs with zero human input [3]. In response, industry groups such as the IFPI and the RIAA are actively pushing for standardized, voluntary AI-content labeling programs to ensure clear delineation across the digital music landscape [1][2].