NVIDIA Sued , Over AI Copyright Infringement.
Authors have sued NVIDIA over NeMo, the company's AI language model that helps to create and train chatbots, Engadget reports.
Authors have sued NVIDIA over NeMo, the company's AI language model that helps to create and train chatbots, Engadget reports.
According to authors Abdi Nazemian,
Brian Keene and Stewart O'Nan, their
books were illegally used to train the AI.
They are seeking a jury trial and want NIVIDIA to pay damages and destroy the dataset that powers NeMo's large language models.
According to the authors, the Books3
dataset copied Bibliotek, a shadow library containing 196,640 pirated books.
In sum, NVIDIA has admitted
training its NeMo Megatron models
on a copy of The Pile dataset. , Via lawsuit against NVIDIA.
Therefore, NVIDIA necessarily also
trained its NeMo Megatron models
on a copy of Books3, because
Books3 is part of The Pile. , Via lawsuit against NVIDIA.
Certain books written by Plaintiffs
are part of Books3— including
the Infringed Works—, Via lawsuit against NVIDIA.
... and thus NVIDIA necessarily trained
its NeMo Megatron models on one or
more copies of the Infringed Works,
thereby directly infringing the
copyrights of the Plaintiffs, Via lawsuit against NVIDIA.
NVIDIA responded to the suit, telling 'The Wall Street Journal,' "we respect the rights of all content creators and believe we created NeMo in full compliance with copyright law.".
OpenAI and Microsoft were hit
with a similar lawsuit last year.