NVIDIA just sued again for alleged copyright infringement, the eighth lawsuit it has faced for its training of its AI models.
Plaintiff S.A. JAMENDO alleges that NVIDIA has used its audio files to train NVIDIA models.
The interesting twist is that Jamendo has partnered with Music Technology Group at Universitat Pompeu Fabra in Barcelona, Spain, to offer a free music dataset called MTG-Jamendo for researchers to use under a Creative Commons license:
The Complaint alleges that NVIDIA went beyond the terms of the Creative Commons license to use the dataset for commercial purposes to train NVIDIA’s models:
Jamendo alleges (1) copyright infringement in the selection and arrangement of the dataset, (2) copyright infringement in the musical works and sound recordings, (3) Breach of Contract – CC BY-NC 4.0 License, (4) Breach of Contract – Terms of Use, (5) Unjust Enrichment, (6) Unfair Competition – Cal. Bus. & Prof. Code § 17200.
Jamendo seeks about $20,269,750.00 USD (17.8 million Euro) in actual damages and NVIDIA profits, or, alternatively, statutory damages.
DOWNLOAD THE COMPLAINT
