This is not the first case of an AI company being sued over copyright infringement related to training data. In December 2021, The New York Times sued Microsoft and OpenAI for allegedly using its articles without permission to build AI models. Nvidia, however, maintains that it created NeMo in full compliance with copyright law. The authors are seeking a jury trial and damages for the alleged copyright violations.
Key takeaways:
- Nvidia is facing a proposed class action lawsuit for allegedly using copyrighted works to train AI models in its Megatron library without the authors' permission.
- The lawsuit was filed by three authors who claim their books were used in the training of the Megatron large language models.
- The models in question were released in September 2022 and were trained on "The Pile" dataset, which is known to contain a number of unlicensed copyrighted works.
- This follows a similar case where The New York Times sued Microsoft and OpenAI for using its articles without permission to build AI models.