Publisher Sues Anthropic and Meta Over AI Training

Carlson LLC alleges the companies used pirated and scanned books without permission to train their language models.

Updated on Nov. 21, 2026 in Artificial Intelligence

Stack of closed hardcover books with uncut page edges in front of dark server cabinets under cool overcast daylight.
Carlson LLC sued Anthropic and Meta in federal court, alleging the companies used pirated and scanned copyrighted books without permission to train their AI language models. AI Illustration. Upload story photo >

Live Poll

Should AI companies pay authors before training models on their copyrighted books?

Carlson LLC filed a federal copyright infringement lawsuit against Anthropic PBC and Meta Platforms, Inc. in the U.S. District Court for the Northern District of California. The complaint alleges the companies harvested training materials from pirate libraries Books3 and Anna's Archive, and acquired books by scanning physical copies without permission.

Why it matters

The case adds to a growing set of legal challenges over whether AI developers can use copyrighted books to train models without licensing them. Carlson LLC opted out of the Bartz class action settlement with Anthropic, so it is pursuing its claims individually.

The complaint names Anthropic PBC and Meta Platforms, Inc. as defendants in a single filing under case number 3:26-cv-14248. The allegations about how training data was acquired remain unproven, and the defendants have not yet responded on the merits.

The players

Anthropic PBC

Anthropic is an artificial intelligence company that develops the Claude family of large language models.

Meta Platforms, Inc.

Meta Platforms is the technology company that owns Facebook, Instagram and WhatsApp and develops the Llama series of open AI models.

The details

Carlson LLC is the registered copyright owner of literary works written by R. and K. Carlson, including R. Carlson's 1997 book Don't Sweat the Small Stuff... and It's All Small Stuff, which R. Carlson, who died in 2006, wrote. The lawsuit alleges both companies used these works to train AI models, and Carlson LLC argues the bestselling books have been directly harmed by competition from large language models.

Timeline

  1. 1997: R. Carlson published Don't Sweat the Small Stuff... and It's All Small Stuff.

  2. 2006: R. Carlson died.

  3. November 19, 2026: Carlson LLC filed suit in the Northern District of California.

Legal Context

Carlson LLC opted out of the Bartz class action settlement between book authors and Anthropic and is instead pursuing an individual copyright claim. The filing tests whether individual publishers can pursue separate claims against multiple AI developers after collective settlements have been reached.

Authors and publishers who opt out of collective settlements may need to pursue individual claims, which can be slower and more costly. Readers of AI products may see changes in how training data is licensed if courts side with rights holders.

The takeaway

Copyright disputes over AI training data are increasingly playing out in individual lawsuits as well as class actions, so creators should track which settlements they are bound by. Opting out of a class settlement can preserve individual claims but shifts cost and risk to the rights holder.

Further reading

For related coverage of how courts are treating AI training data, see Artificial Intelligence.

Source note: This article includes information reported by Vitallaw.

Live Poll

Should AI companies pay authors before training models on their copyrighted books?