Suno's AI training exposed as it scrapes millions of songs from YouTube
entertainment
controversial
impactful

Suno's AI training exposed as it scrapes millions of songs from YouTube

10
(Update: )
Italian comune
video-sharing platform owned by Google
  • A hacker accessed Suno's source code, revealing its AI was trained on music from platforms like YouTube and Deezer.
  • The breach supports claims from artists and record labels that their music is used without consent.
  • Suno's legal battles highlight the ongoing conflict between AI technology and copyright protections.
Share opinion
1

Story

In late 2025, a significant breach occurred involving Suno, an AI company known for creating AI-generated audio clips and songs. The hacker, identified by the alias ellie.191, executed a supply chain attack that allowed access to Suno's source code. This breach revealed that Suno's AI was trained using a vast collection of music and lyrics sourced from various platforms, including YouTube Music, Deezer, Genius, and stock music libraries. The investigation into the breach confirmed that over 2 million audio clips were stored in a folder labeled 'youtube_music', alongside extensive hours of music from other sources. The implications of this hack are profound, as it substantiates long-standing claims made by musicians, streaming services, and record labels. These entities have accused AI music services of utilizing their copyrighted material without consent or proper licensing. The ongoing legal battles highlight the tension between traditional music rights and the emerging AI music industry. Suno, like other AI companies, argues that its use of original material for training purposes falls under the fair use doctrine in copyright law, a claim that has been contested in courts. In response to the breach, Suno's spokesperson stated that the compromised source code was outdated and that no sensitive personal information was leaked. They also emphasized that they do not have access to customers' full credit card numbers, asserting compliance with current privacy laws. Despite these reassurances, the breach raises questions about the ethical implications of AI training methods and the potential for copyright infringement. As the popularity of AI-generated music continues to rise, the situation underscores the need for clearer regulations and agreements between AI companies and content creators. Some publishers and copyright holders have begun to negotiate deals with AI firms to allow the use of their content, but many remain wary of the potential for exploitation. The legal landscape surrounding AI-generated music is evolving, and the outcomes of these lawsuits could set important precedents for the future of both the music industry and AI technology.