Hack Reveals Suno AI Scraped Millions of Songs from YouTube, Deezer, and Other Platforms
Key Takeaways
- ▸Hacked data confirms Suno scraped millions of copyrighted songs from YouTube Music, Deezer, Genius, and multiple other music platforms and libraries
- ▸The breached source code reveals Suno ingested over 600,000 hours of music content, demonstrating the massive scale of unauthorized data collection
- ▸The hack exposed user data for hundreds of thousands of customers and Stripe payment information, raising serious privacy and security concerns
Summary
A security breach of Suno, one of the largest AI music generation tools, has exposed details about the company's training data sources, revealing that the company scraped millions of songs and lyrics from YouTube Music, Deezer, Genius, Pond5, Jamendo, Freesound, the International Music Score Library Project, and podcasts via RSS feeds. The hacked data includes source code from 2023-2024 showing scraping instructions and dataset details, indicating Suno ingested over 600,000 hours of music content, with YouTube Music alone contributing over 113,000 hours. The breach also exposed user information for hundreds of thousands of Suno customers and Stripe payment information.
The revelation provides concrete evidence supporting claims by the Recording Industry Association of America and multiple record labels that have sued Suno for copyright infringement. While Suno has previously argued it trained on "essentially all music files of reasonable quality that are accessible on the open internet" as an exercise of fair use, the hacked data now reveals the specific mechanisms and scope of this data collection. The source code shows Suno used scraping tools from Bright Data to access content through proxies and specifically targeted acapella versions of songs on YouTube.
- Evidence from the breach directly supports ongoing copyright lawsuits from the Recording Industry Association and record labels
