Mark Zuckerberg’s Meta allegedly trained AI on millions of stolen books from “shadow libraries”

04/06/2025 / By Laura Harris

Meta allegedly trained its AI model, Llama 3, using millions of pirated books from shadow libraries like LibGen and Anna’s Archive without compensating authors or publishers.
Court documents show Meta executives prioritized books over web data, with one email stating, “Books are actually more important than web data.”
Works by prominent Australian writers, including Sophie Cunningham, Hannah Kent and Tim Winton, were found in LibGen’s database, allegedly used by Meta for AI development.
Writers expressed anger over unauthorized use, with Cunningham calling it exploitative and Kent describing it as “plundering” her work. Both are exploring legal measures.
Critics argue Meta’s actions highlight unethical AI training practices, exploiting copyrighted material despite its $164B revenue, raising concerns about fair compensation for creators.

Meta CEO Mark Zuckerberg is facing new allegations that his company deliberately used millions of pirated books from notorious shadow libraries, including LibGen and Anna’s Archive, to train its advanced artificial intelligence (AI) model, Llama 3, without compensating authors or publishers. (Related: Ireland fines Meta $102M for lapse in PASSWORD SECURITY.)

According to court documents, Meta executives openly discussed their need for high-quality literary content, with one damning email stating, “Books are actually more important than web data.” Rather than licensing material legally, the company allegedly turned to piracy sites hosting vast troves of stolen works, including academic papers and novels.

On March 20, U.S. magazine The Atlantic published a tool allowing the public to search LibGen’s database for the first time. The investigation identified works by prominent Australian authors – including Charlotte Wood, Alexis Wright, Tim Winton, Helen Garner and Sophie Cunningham – among the materials allegedly used by Meta to develop its AI systems.

LibGen, a digital repository with over 7.5 million pirated books and 81 million research papers, was reportedly a key source for the data scraping of Meta, the parent company of Facebook and Instagram. The same was true for Anna’s Archive, another infamous hub for copyrighted material.

All this, despite Meta’s 2024 staggering revenues exceeding to $164 billion and profits approaching to $62 billion.

With such vast resources at its disposal, the company undoubtedly could compensate creators, publishers and researchers fairly. Yet, Meta allegedly opted to exploit their content for AI training without permission.

Australian authors outraged as Meta uses their work to train AI without consent

Some of Australia’s most celebrated authors are furious after discovering that Meta has used their copyrighted works to develop its AI platform without their consent.

Prominent author Sophie Cunningham, whose latest novel, “This Devastating Fever” (2022) was found in LibGen’s database, expressed her anger in an interview with ABC Arts.

“I’m really angry about it,” Cunningham said. “The average writer earns about $18,000 a year on their writing. It’s one thing to be underpaid. It’s another thing to find that work is being used by a company that you don’t trust.” Cunningham is now considering legal action and has asked her publishers to issue cease and desist notices to Meta on her behalf.

Bestselling author Hannah Kent was equally shocked to find all three of her novels in LibGen. “I felt completely gutted,” Kent said. “It feels a little like my body of work has been plundered.”

“As a writer, what I would like, at the very least, is for someone to seek permission,” Kent, whose memoir, “Always Home, Always Homesick,” to come out this April, said. “This is what happens when someone wants to reproduce my work in any other way … even if it is for fair use.”

Moreover, Kent fears the implications this could have for the tech industry. “The fact that we have Meta not only training AI and trying to claim that under fair use but also using a shadow library of pirated material indicates the lack of ethical consideration that is being put into training AI. It opens the door to others also feeling like this is an acceptable way to treat intellectual copyright and creatives who already are expected to [contribute] so much for free or without due recompense.”

Visit MetaTyranny.com for similar stories.

Watch this video of Google and Big Pharma being exposed through a Meta-AI analysis.

This video is from the Haixus channel on Brighteon.com.