Music-tech firm MatchTune is weighing into the debate around musical AIs and training, with a new tool that it says can identify tracks made using the Suno platform.
“We’ve successfully cracked the Suno music generation model and developed an advanced detection technology that identifies SUNO-generated content with an impressive 90% accuracy (0.01% false positives and 90% false negatives),” wrote MatchTune’s chief business development and rights officer Virginie Berger in an email to industry contacts.
“This breakthrough represents a huge leap in safeguarding artists’ work from unauthorized AI-generated replications… Our team meticulously dissected the AI music generation process, identifying unique Suno signatures and patterns. This allows us to recognize Suno content in less than a second, enabling rapid scanning of vast music catalogs.”
She added that MatchTune is now working on cracking the systems of Suno rivals Udio and So-Vits – the latter being an open-source project used for voice cloning.
