Tech Policy at Duke University

On March 17, Grammy-nominated artist Tift Merritt and renowned guitarist and composer Marc Ribot came together at Duke’s Periodic Tables, hosted by Science & Society, to discuss the threats that AI-generated music poses to musicians today. Both artists have spent years advocating for musicians’ rights–Ribot as a member of the Music Workers Alliance and Merritt through her work with the Human Artistry Campaign. While advocacy for this, often underrepresented, group of creatives has always been important, the rapid rise of AI-generated music makes ensuring their voices are heard more urgent than ever.

The event opened with the playing of a song that sounded eerily like Tift Merritt’s voice, yet something felt slightly off. The host of Periodic Tables, Professor Misha Angrist, revealed that the track was generated by Suno AI. Prof. Angrist prompted Suno AI to create a song in Merritt’s style, which it did without hesitation. The demonstration highlighted the risks that AI-generated music poses to musicians’ livelihoods, raising urgent questions around consent and copyright infringement.

Marc Ribot responded to the song by quoting Suno AI CEO Mikey Shulman, who claimed, “It’s not really enjoyable to make music anymore.” Ribot highlighted the irony of this assertion: the head of an AI company (someone with no firsthand experience as a professional musician) making sweeping claims about how musicians feel about their craft. His response emphasized the disconnect between those building AI systems and the artists on whose work such systems depend.

Both speakers acknowledged the benefits that AI can bring to music-making. Ribot even noted that he has used it in processes like track cleaning. However, he stressed the need to draw a clear line, explaining, “what I put into AI was my own tracks.” In his view, using AI systems trained on other musicians’ work without consent crosses into exploitation. The message was clear: artists are not opposed to technological innovation, but they should have agency over how and when it is used in their work.

The broader issue, they argued, is the underrepresentation of musicians in the development of technologies that rely on their creative output. As Merritt put it, “it’s imperative that artists are at the table talking to tech [companies].” Without inclusion, the systems being created today risk devaluing artist labor rather than supporting it. Artists should consent to how their work is used, particularly in the training of AI models, to ensure creative output is not exploited without permission or compensation.

The ongoing advocacy efforts of Marc Ribot and Tift Merritt have the potential to shape the future of the industry and improve conditions for working musicians. As Merritt warned, “tech inserting itself into an industry without talking to that industry, deflating the wages…in order to extract corporate profits is definitely a business model and it’s not going to stop with musicians.” She frames musicians as the unfortunate “canary in the coal mine,” signaling broader risks for other creative and labor sectors as AI continues to expand. Marc Ribot and Tift Merritt’s work to push for greater transparency and better artists’ rights highlights a potential path toward a more equitable relationship between musicians and emerging technologies.

More from the AI & the Creative Economy Project

Currently, independent musicians cannot collectively bargain due to a ruling from the 1980s that classifies musicians as independent contractors. The Protect Working Musicians Act (PWMA) offers a practical, market-based solution by allowing independent music creators to collectively bargain with dominant online platforms where they do not currently have a place at the table. Sign the Protect Working Musicians Act Petition!

Learn more about the AI & Creative Economy Project and Periodic Tables at Duke.