Nvidia sued for training NeMo on authors' copyrighted works

Occurred: March 2024

GPU chip provider Nvidia has been sued by 3 authors accusing it of training it's NeMo models on copyrighted books.

Nvidia's NeMo platform is a tool for enterprises to develop custom GenAI. In October 2023, Nvidia acknowledged training Nemo's models on a dataset containing approximately 196,640 books and removed the set amidst copyright infringement concerns. The Books3 dataset contains the same number of books.

In March 2024, authors Brian Keene, Abdi Nazemian and Stewart O'Nan submitted a class action lawsuit against Nvidia for copyright infringement, saying their works were part of this dataset and were trained on without their permission. They are seeking unspecified damages.

System 🤖

Operator: Nvidia
Developer: Nvidia
Country: USA
Sector: Media/entertainment/sports/arts
Purpose: Train and deploy custom LLMs
Technology: Generative AI; Machine learning; Neural network; Deep learning; NLP/text analysis
Issue: Copyright
Transparency: