Authors Sue Nvidia for Copyright Infringement in AI Training Platform NeMo

Tech giant Nvidia is facing a lawsuit from a group of authors who claim that their copyrighted works were used without permission to train its artificial intelligence platform NeMo. The authors, Brian Keene, Abdi Nazemian, and Stewart O’Nan, allege that their works were included in a dataset of 196,640 books that were used to train NeMo before being removed in October due to reported copyright infringement.

The proposed class action lawsuit, filed in San Francisco federal court, accuses Nvidia of admitting to training NeMo on the dataset, thereby infringing on the authors’ copyrights. This lawsuit is part of a larger trend of legal action against tech companies for AI copyright infringement.

The authors seek unspecified damages for individuals in the U.S. whose copyrighted works were used to train NeMo’s large language models in the last three years. These models are essential for powering AI tools like NeMo, which Nvidia touts as a fast and affordable way to adopt generative AI.

Among the works included in the lawsuit are Keene’s “Ghost Walk,” Nazemian’s “Like a Love Story,” and O’Nan’s “Last Night at the Lobster.” The lawsuit claims that these books were part of a dataset known as “The Pile,” which Nvidia used to train its NeMo Megatron AI models.

Nvidia’s stock has surged over the past year, making it a leading AI chipmaker with a market value of nearly $2.2 trillion. However, the company now faces legal challenges over the alleged use of copyrighted content in its AI models.

This lawsuit adds Nvidia to a growing list of tech companies facing legal action over AI copyright infringement, including high-profile cases involving the New York Times, OpenAI, and Microsoft. As the use of AI continues to expand, the issue of copyright infringement in training datasets is likely to remain a contentious issue in the tech industry.