NVIDIA is being sued by three authors who allege that the company behind the H100 and other highly desirable AI GPUs had used their copyrighted books without permission to train language models for NeMo—one of its AI platforms—over the past three years, Reuters has learned. NeMo is described by NVIDIA as an “end-to-end, cloud-native framework” that allows customers to build, customize, and deploy generative AI models anywhere.
Authors behind the lawsuit:
- Abdi Nazemian
- Brian Keene
- Stewart O’Nan
Books mentioned include:
- Last Night at the Lobster (2007)
- Perched in the far corner of a run-down New England mall, the Red Lobster hasn’t been making its numbers and headquarters has pulled the plug. But manager Manny DeLeon still needs to navigate a tricky last shift–just four days before Christmas and in the midst of a fierce blizzard–with a near-mutinous staff and the final onslaught of hungry retirees, lunatics, and holiday office parties. All the while, he’s wondering how to handle the waitress he’s still in love with, his pregnant girlfriend, and where to find the present that will make everything better.
- Ghost Walk (2008)
- Halloween is coming, and a new haunted attraction is open for business in LeHorn’s Hollow. Folks will come from miles around to walk down the spooky trail and get scared witless. But there’s one thing the owners of the ghost walk haven’t counted on. There really is something waiting in the woods-a vast, ancient evil whose hunger threatens to consume all life on Earth. Soon, the unsuspecting customers will pay their money and get in line… to die. And only one man, occult detective Levi Stoltzfus, can save them and stop the evil before it is unleashed.
- Like a Love Story (2019)
- Reza is an Iranian boy who has just moved to the city with his mother to live with his stepfather and stepbrother. He’s terrified that someone will guess the truth he can barely acknowledge about himself. Reza knows he’s gay, but all he knows of gay life are the media’s images of men dying of AIDS.
NVIDIA has reportedly already admitted guilt:
Brian Keene, Abdi Nazemian and Stewart O’Nan said their works were part of a dataset of about 196,640 books that helped train NeMo to simulate ordinary written language, before being taken down in October “due to reported copyright infringement.”
In a proposed class action filed on Friday night in San Francisco federal court, the authors said the takedown reflects Nvidia’s having “admitted” it trained NeMo on the dataset, and thereby infringed their copyrights.