Generative audio platform Suno is challenging the validity of artist catalogs for AI training as Universal Music and Sony Music pursue legal action against the company, musicbiz.rockol.it reported.
The core dispute centers on whether recording agreements permit the extraction and training of machine learning models. Record labels argue that existing copyright protections cover these modern computational uses.
Suno’s defense attacks the fundamental assumption that legacy agreements ever contemplated neural network training. Companies built on transformer architectures and latent diffusion models now find themselves locked in courtroom battles over intellectual property rights.
Recording agreements fail to address modern computational audio generation. Legal teams representing the technology firm point to the age of these contracts to undermine the plaintiffs’ licensing claims.
Sono Music agreements do not provide for AI, Britt Lovejoy of Suno’s legal team explained to musicbiz.rockol.it.
“As one would expect, artist licensing contracts—some of which, in the case of Sony Music, date back to the 1980s—do not provide for AI,” Britt Lovejoy told musicbiz.rockol.it.
“They do not explicitly state whether Sony Music…,” Lovejoy added, explaining the contractual ambiguity surrounding legacy catalogs.
The outcome of this litigation will determine how platform developers negotiate training permissions with major labels moving forward. Current negotiations rely heavily on the assumption that rights holders control clear pathways for licensing historical repertoires to tech firms.
Legal teams on both sides will continue filing briefs regarding contractual interpretation as the court reviews the scope of agreements.
