Nvidia faces lawsuit from three authors over AI training practices

CALIFORNIA, UNITED STATES — Nvidia Corporation joins the growing list of tech giants facing legal battles over generative artificial (AI) copyright issues.

The $2.2 trillion company, whose high-powered chips drive AI, is facing a class action lawsuit from three authors alleging copyright infringement. 

Brian Keene, Abdi Nazemian, and Stewart O’Nan claim Nvidia unlawfully used their copyrighted books to train its NeMo language model without permission.

The authors state their works, including Keene’s “Ghost Walk,” Nazemian’s “Like a Love Story” and O’Nan’s “Last Night at the Lobster,” were part of a 196,640 book dataset used to enhance NeMo’s language capabilities before being removed last October, Reuters reported.

Filed in a San Francisco federal court, the lawsuit seeks compensation for all U.S. writers whose materials were used to train Nvidia’s AI over the past three years. 

Nvidia declined to comment on the legal matter. The new case adds to growing legal scrutiny around AI training methods involving copyrighted text, images, video, or audio.

Aside from Nvidia, Microsoft and OpenAI were also sued by The New York Times and other authors over the same issue.

Meanwhile, a new nonprofit organization called Fairly Trained aims to resolve simmering tensions over copyright infringement in the AI sector. It offers certifications to AI companies that train models only on data from consenting creators.

Nine companies have already gained Fairly Trained certifications, including generative audio startups Endel and LifeScore.