Skip to content
Tech News
← Back to articles

Spotify is deleting millions of AI-generated music tracks to fend off spammers

read original more articles
Why This Matters

Spotify's efforts to remove millions of AI-generated tracks highlight the ongoing challenge of combating AI-driven spam and fake content on digital platforms. This underscores the need for robust moderation and the evolving role of AI in content creation, which impacts both industry integrity and consumer trust. As AI becomes more integrated into music production, distinguishing authentic art from AI-generated spam remains a critical concern for the industry.

Key Takeaways

SpotiSlop: Large language models and sophisticated audio-generation tools are disrupting the traditional music industry. Digital listening platforms such as Spotify have become prime targets for AI-powered spammers, but trade organizations are fighting back with new labeling programs.

Spotify's Sam Duboff recently revealed that the platform was forced to remove 75 million AI-generated tracks in 2025 alone. As Spotify's senior director and global head of marketing, policy, and music business, Duboff believes that AI has not introduced any entirely new tactics for spam operations. However, he also argues that generative AI and other machine learning technologies have taken audio spam to the next level.

Spotify now has systems designed to combat both AI-generated "slop" uploads and data scraping by companies seeking new content to train their generative AI models. Duboff confirmed that the company has a "big team" dedicated to identifying potential new attack vectors that could make life easier for spammers (or scraping bots).

The executive also provided several eye-opening figures highlighting the growing prevalence of AI-generated music on digital platforms. Every day, bots upload around 100,000 different "songs" to Spotify's servers, and a large portion of these tracks are most likely "made" using generative AI services or custom LLM-based setups.

Duboff acknowledges that not every genAI track can simply be dismissed as slop. The 75 million songs removed last year were low-effort content with very little human creativity involved beyond a lazy chatbot prompt. However, "real" artists are increasingly using AI technology in their production workflows, blurring the lines between fake music and human-curated efforts.

Spotify is certainly embracing AI across its business these days. Earlier this year, co-CEO Gustav Söderström said that the company's engineers are essentially allowing AI agents to handle much of their coding work. Spotify also said that AI tools approved by music labels are now fair game for creating remixes and covers that can then be sold on the platform.

Beyond Spotify's growing pains with spammy tracks, generative AI is forcing the entire music industry to develop new solutions to the increasingly relevant AI slop problem. The International Federation of the Phonographic Industry, the RIAA, and other major trade organizations recently announced a "voluntary" program to properly label AI-generated music, giving listeners a clear indication when a song has been entirely created through a chatbot prompt. These labels should also identify "AI-assisted" music that was primarily created by human artists.