Authors Sue Nvidia Over Unauthorized Use of Copyrighted Works in AI Training
Nvidia, renowned for its AI-enabled chips, faces a lawsuit from three authors who allege that the company used their copyrighted books without authorization to train its NeMo AI platform.
Authors Brian Keene, Abdi Nazemian, and Stewart O’Nan claim that their works were included in a dataset of approximately 196,640 books used to train NeMo in simulating natural written language. The dataset was reportedly taken down in October due to copyright infringement concerns.
In the lawsuit filed in San Francisco federal court on Friday night, the authors argue that Nvidia’s acknowledgement of training NeMo on the dataset constitutes copyright infringement. They seek unspecified damages on behalf of U.S. copyright holders whose works contributed to training NeMo’s large language models over the past three years.
The lawsuit mentions Keene’s “Ghost Walk” (2008), Nazemian’s “Like a Love Story” (2019), and O’Nan’s “Last Night at the Lobster” (2007) as among the works included without permission.
Nvidia declined to comment on the matter, while legal representatives for the authors have yet to respond to requests for additional statements.
This legal action adds Nvidia to a growing list of companies facing litigation from writers and organizations like the New York Times concerning generative AI, which generates new content based on various inputs such as text and images.
NeMo is touted by Nvidia as an efficient and cost-effective solution for adopting generative AI technology.
Several other companies, including OpenAI and Microsoft, have also faced lawsuits related to generative AI technology.
Nvidia’s prominence in the AI sector has attracted substantial investor interest, with its stock price surging nearly 600% since the end of 2022, resulting in a market value of nearly $2.2 trillion.
The case, identified as Nazemian et al v Nvidia Corp, is being heard in the U.S. District Court for the Northern District of California, with case number 24-01454.
