AI music startup Suno argues that training on copyrighted music constitutes 'Fair Use'
AI music startup Suno asserts that training its model on copyrighted music qualifies as 'fair use,' sparking debate over AI and intellectual property rights.
In response to the recent lawsuit filed by the Recording Industry Association of America (RIAA) against music generation startups Udio and Suno, Suno has admitted in a court filing that it trained its AI model using copyrighted songs. However, the company argues that this practice is protected under the fair-use doctrine.
The RIAA's lawsuit, filed on June 24, accuses Udio and Suno of training their models with copyrighted music without permission. Although Suno's investors had previously hinted that the startup might not have had authorization from music labels, this was explicitly confirmed in the latest filing.
According to the court documents, “It is no secret that the tens of millions of recordings that Suno’s model was trained on presumably included recordings whose rights are owned by the Plaintiffs in this case.”
Suno CEO and co-founder Mikey Shulman elaborated in a blog post published alongside the legal filing, explaining that the company trains its models using medium- and high-quality music available on the open internet. Shulman compared this process to a musician learning from listening to music, asserting that "Learning is not infringing. It never has been, and it is not now."
The RIAA responded sharply, accusing Suno of attempting to downplay their infringement. They stated, “This is a major concession of facts they spent months trying to hide and only acknowledged when forced by a lawsuit. Their large-scale infringement does not qualify as ‘fair use’. There’s nothing fair about stealing an artist’s life’s work, extracting its core value, and repackaging it to compete directly with the originals… Their vision of the ‘future of music’ is one where fans may no longer enjoy music from their favorite artists because those artists can no longer earn a living.”
As the legal battle continues, the case highlights ongoing tensions between emerging AI technologies and established intellectual property rights, raising important questions about the boundaries of fair use in the digital age.
The issue of fair use has always been complex, and with the advent of AI model training, even established doctrines may face challenges. The outcome of this case, still in its early stages, is poised to set a significant precedent that could impact not only the two startups involved but also shape the future landscape of intellectual property and AI development.