The ongoing legal fight over using publishing content to train large language models has taken a new turn.
Entrepreneur Media LLC, the publisher
of Entrepreneur Magazine and many books, filed suit against Meta Platforms Inc. on Thursday, alleging that the social media firm “pirated at least hundreds of copyrighted works
owned by Entrepreneur,” and that it obtained these works from “notorious ‘shadow libraries.'”
The 16-page complaint charges that Meta
“deliberately and willfully stole hundreds of terabytes of copyrighted works to train its family of LLMs, known collectively as Llama."
Meta has invested billions of dollars in
developing Llama, but not in paying Entrepreneur for using its content, it adds.
This alleged theft of content involved using “torrenting” sites, the
complaint continues.
advertisement
advertisement
It states that “Meta utilized torrenting sites such as BitTorrent and LibTorrent to download and redistribute Entrepreneur’s
protected works from shadow libraries, including LibGen. By using these protocols, Meta not only reproduced Entrepreneur’s works but also automatically distributed them to other users in the
torrenting network, compounding the infringement and facilitating global piracy.”
The complaint continues, “Upon information and belief, Meta
developed and deployed software designed to process pirated books for training, including functionality to remove copyright management information (“CMI”) such as copyright notices,
metadata, and author identifiers. Meta did so to conceal its infringement and to facilitate further unauthorized use.”
Many copyright lawsuits are directed
against firms like OpenAI. But the Entrepreneur complaint notes that “companies including Anthropic, OpenAI, and Google have paid billions of dollars to obtain proactive and
retroactive licenses for LLM training—and according to public records, Meta itself has AI-licensing deals with Reuters, Shutterstock, and Universal Music Group."
The case was first reported by Reuters.
The suit is on file with the US. District Court for the Northern District of California, San Francisco
division.