Music Industry Deploys AI Detection To Combat Streaming Fraud

Music streaming platform Deezer is implementing new measures to combat the rise of AI-generated songs used to commit streaming fraud, including labeling albums with AI-created tracks and excluding such songs from editorial playlists and recommendations. Deezer reports that about 18% of daily uploads, roughly 20,000 songs, are fully AI-generated, with an estimated 70% of streams on these tracks being fraudulent to claim royalties illegitimately. This surge in AI music challenges the industry’s existing structures, prompting the adoption of detection systems embedded throughout the music pipeline—from training datasets to distribution platforms—to trace and tag synthetic content early. Startups and companies like Vermillio and Audible Magic are developing infrastructure with tools such as TraceID to authenticate releases and replace older systems like YouTube’s Content ID, aiming to protect artists’ rights and ensure proper credit and payments. While AI offers creative benefits, the music industry is focused on transparency and governance to prevent abuse and maintain fairness for human creators. Deezer’s efforts highlight an industry-wide issue of generative AI disrupting music streaming and royalty systems, necessitating proactive detection and labeling strategies.