Last month, The Atlantic’s AI Watchdog uncovered a massive dataset of over 12 million songs allegedly being used to train AI. Now, hackers have discovered a source code in Suno AI that potentially exposes a grand-scale act of data scraping across decades of copyrighted music.

According to 404 Media, hacker ellie.191 discovered that Suno had scraped millions of hours of songs, lyrics and podcasts from Deezer, Genius, YouTube Music, and a slew of stock music libraries. Though, it might not come as too much of a shock; the company infamously insisted it was “fair use” to train its AI model on copyrighted songs back in 2024.

• READ MORE: “The magic, consistency, and mystique of Rick Rubin really started to fall off”: music critic Anthony Fantano thinks producer Rick Rubin is ‘cringe’

While Suno has faced multiple lawsuits for “trampling” on artist’s rights, the exposed code seems to back up the June 2024 joint lawsuit pieced together by Universal Music Group, Sony Music Entertainment, Warner Music Group and the Recording Industry Association of America. The lawsuit alleged Suno and Udio were carrying out “unimaginable scales” of copyright infringement due to stream-ripping from Youtube – and the code seems to suggest exactly that.

In screenshots of code shared by 404 Media, it can be seen that thousands, if not millions, of hours of music from were being ripped from millions of clips, all taken from a number of online platforms. One section of code literally lists that it is taking from “genius_hq, youtube_music, freesound, jamendo, imp, deezer, ytm_tagged”.

Elsewhere, sections suggested that Suno’s coding was also seeking out vocal, acapella videos in particular. Of course, these clips would aid any training for filtering vocals, as well as outright generating entirely AI-voices. This would also be aided by the podcast-centric scrapping, with the leaked code suggesting one million hours of podcasts had been downloaded.

As the hacker tells 404 Media, they uncovered the code by “hacking an individual employee using the Shai-Hulud worm, a supply chain attack that allowed hackers to harvest GitHub and cloud service credentials”. They also found that they were able to access the payment information and personal data of “hundreds of thousands” of Suno users.

Despite the hacker’s access to bank details, phone numbers and addresses, they insist they have no ill intentions. “I like to hack anything and everything,” they say.

The hacked code comes as another blow to the countless artists speaking out against AI. SZA has been particularly vocal about her disapproval of AI as of late, even taking to her private Instagram to condemn Diplo for his support of AI tools. “If you’re a musician and you support this degenerate shit, you’re disgusting and there’s nothing you could ever say to me to make this okay,” she wrote.

View this post on Instagram

The post Hackers reveal Suno scraped music data from YouTube, Deezer, Genius, and more appeared first on MusicTech.