Archive · 22 Jan 2026
Nvidia Approved Use of Pirated Books for AI Training, Lawsuit Alleges
Nvidia is in trouble because some authors say the company used books that were stolen from the internet to teach its AI.
The authors are suing Nvidia.
Some documents show that Nvidia’s leaders might have known about this and even agreed to pay for access to these stolen books.
This is important because using someone else’s work without permission is not allowed, and it’s a big deal in the world of technology and law.
Nvidia allegedly used pirated books from Anna’s Archive and Bibliotik to train its AI models.
The company’s data strategy team reportedly offered to pay for high-speed access to these shadow libraries.
Nvidia’s top leadership allegedly approved the payment plan within a week.
The allegations are part of an ongoing class action lawsuit by authors accusing Nvidia of copyright infringement.
The lawsuit involves the Books3 dataset, which includes copyrighted works from pirate sites.
- Who
- Nvidia and authors involved in a class action lawsuit
- What
- Alleged use of pirated books for AI training
- When
- Approval process occurred within a week
- Where
- Anna’s Archive, Bibliotik, and Books3 dataset
- Why
- To train AI models, allegedly approved by Nvidia’s top leadership
- Company Involved
- Nvidia
- Alleged Action
- Use of pirated books for AI training
- Source of Alleged Pirated Content
- Anna’s Archive, Bibliotik
- Dataset Involved
- Books3
- Legal Action
- Class action lawsuit by authors
- Approval Process
- Allegedly approved by Nvidia’s top leadership
- Document Source
- TorrentFreak, court filings