Publishers sue ‘shadow’ library allegedly powering AI chatbots

6 March 2026

(From Reuters)

March 6 (Reuters) – A group of publishers including the “Big Five” English-language book publishing houses — Hachette, Penguin Random House, HarperCollins, Macmillan and Simon and Schuster — filed a lawsuit in New York federal ​court on Friday seeking to shut down prominent “shadow library” Anna’s Archive.

The lawsuit alleged that ‌Anna’s Archive has pirated over 63 million books and provides them to companies for artificial intelligence training.

“Anna’s Archive is a brazen pirate operation that steals and distributes millions of literary works while outrageously offering access to AI ​developers in exchange for crypto payments,” Maria Pallante, president of the publishers’ trade group the ​Association of American Publishers, said in a statement.

Spotify and a group of ⁠major music publishers filed a related New York lawsuit against Anna’s Archive in January seeking trillions ​of dollars in damages for pirating tens of millions of audio files. The site, which is ​run by unknown individuals, has not responded to that case.

The book publishers’ lawsuit said that the site has continued its piracy in defiance of an early court order in the audio piracy case.

Federal courts can order U.S. internet ​service providers to stop hosting pirate sites. The publishers asked the Manhattan court to shut ​Anna’s Archive down and requested damages of up to $150,000 per pirated work.

While it is difficult for U.S. courts to ‌fully ⁠disable websites based overseas, Pallante told Reuters in an interview that she is hopeful that the lawsuit will at least disrupt Anna’s Archive’s operations.

“It comes down to doing nothing or doing what we can,” Pallante said.

Friday’s lawsuit — which also includes academic publishers Cengage, Elsevier, McGraw Hill and John Wiley & ​Sons as plaintiffs — said ​the self-proclaimed “world’s largest shadow ⁠library” launched in 2022 and allows users to download pirated material.

The publishers said Anna’s Archive claims to have given “high-speed access” to the stolen works ​to customers in China, Russia and other countries for the training of ​AI large ⁠language models that power chatbots and other systems.

Several groups of copyright owners including authors, artists and music companies have filed copyright lawsuits against tech companies for allegedly misusing their work to train AI systems.

The ⁠lawsuits ​say companies like OpenAI and Anthropic pirated their works from ​similar shadow libraries. Lawsuits against Meta and Nvidia have specifically accused the tech giants of downloading books from Anna’s Archive ​to train large language models.

The defendants in those cases have denied wrongdoing.

Reporting by Blake Brittain in Washington