Authors Challenge Nvidia’s AI Training with Copyright Infringement Claim

Authors Challenge Nvidia’s AI Training with Copyright Infringement Claim

Nvidia, a major player in AI chips, is facing a copyright lawsuit from three authors: Brian Keene, Abdi Nazemian, and Stewart O’Nan. The lawsuit alleges that Nvidia used the authors’ copyrighted books without permission to train its NeMo large language model platform.

The crux of the case revolves around a dataset of 196,640 books used to train NeMo. The authors claim their books were included in this dataset, which Nvidia supposedly used to help NeMo mimic real-world written language. Notably, the NeMo platform was shut down in October 2023 amidst copyright infringement concerns. The authors view this removal as an implicit admission of wrongdoing by Nvidia.

This lawsuit seeks unspecified damages for US copyright holders whose works were allegedly used in training NeMo over the past three years. It’s filed as a class action, potentially representing a wider group of affected authors.

This case adds Nvidia to a growing list of tech companies facing legal challenges related to generative AI, a technology that creates new content based on existing data. The New York Times has also filed a similar lawsuit, and companies like OpenAI (creators of ChatGPT) and Microsoft are facing related legal issues.

The lawsuit highlights the complex legal questions surrounding copyright and the training of AI models.