Hack Reveals Suno AI Music Generator Scraped YouTube, Deezer, and Genius

A hacker breached Suno and shared stolen source code and user data with 404 Media. The breach affected hundreds of thousands of Suno customers and exposed Stripe payment data.

The leaked source code (dating from 2023–2024) contained scraping instructions that confirmed Suno's training data sources. YouTube Music was the largest: over 2 million clips totaling roughly 113,879 hours of audio. Deezer contributed about 12,287 hours, and Genius ("genius_hq") accounted for about 17,615 hours. Other sources included Pond5 (62,117 hours), the International Music Score Library Project (19,514 hours), Jamendo (3,726 hours), Freesound (410 hours), and Musescore lyrics (103 hours). Podcasts were also ingested via RSS feeds. In total, the documented datasets amount to at least decades' worth of music.

A code comment noted the pipeline would pull from multiple sources with a step to filter out non-music content.

Suno faces several major record-industry lawsuits. The company previously acknowledged training on "essentially all music files of reasonable quality that are accessible on the open internet" and has argued this constitutes fair use. One suit has already been settled. The article notes the hacked data provides a rare look at exactly how AI music models are built.