SUNO TRAINED ON MILLIONS OF SCRAPED SONGS

Suno, an AI music generation platform, scraped millions of songs and lyrics from major streaming services and lyric databases to train its models, according to data exposed in a hacking incident and reported by 404 Media. The leaked information shows the company extracted content from YouTube Music, Deezer, and Genius—platforms it had not previously disclosed as sources. Suno has maintained opacity around its training datasets and how they were obtained. The revelation highlights a growing tension in AI development: the distinction between fair use and unauthorized data collection. Music industry stakeholders have increasingly challenged AI companies over training practices, with artists and rights holders arguing that large-scale scraping without consent or compensation violates intellectual property protections. Suno has faced legal scrutiny from major record labels, including RIAA members, over its training methodology. The company has previously defended its approach as legal but provided limited public detail about its data sourcing. This leak provides concrete evidence of practices the company had largely kept private.