Bartz v. Anthropic

 by Sybil Johnson

I’m sure a lot of you have heard of the Bartz v. Anthropic lawsuit. Of the dozens of lawsuits about unauthorized use of works for AI training, this is the only one so far that has reached a settlement. Sort of. I’ll get to that later, but first, here are the basics. 

Before I continue let me note that I am not a lawyer nor do I know much about the law. So I’m not an expert in this. This is my take on the subject.

Awhile back several authors sued Anthropic over the use of their works for AI training. From Writer Beware: “...AI company Anthropic over its creation of an enormous library of digitized books to train its Claude LLMs. In addition to purchasing and scanning physical books (a la Google for its Google Books project), Anthropic also downloaded thousands of books that had been illegally uploaded to pirate sites, copying them multiple times for use in AI training.” This became a class action suit at some point in time.

Getting back to my “sort of” comment. There are two parts to this lawsuit. One is the unauthorized use of works for AI training and one has to do with copyright infringement. The judge threw out the AI training portion, deeming it fair use. (I don’t really agree with that.) The other part was settled by Anthropic, a $1.5B settlement. That comes out to about $3,000 per book. That is split between the author and the publisher. If the author publishes their own books, they get the entire amount. The authors who started the suit and, of course, the lawyers get more. After all litigation is closed, Anthropic will destroy the pirated databases LibGen and PiLiMi. Current estimate on finalization is April 2026 with no payouts until then.

The two databases have about 7 million books in them. Only 500,000 are included in the settlement. That’s because only those books that have an ISBN or ASIN and have been registered with the US Copyright office qualify. The registration must have occurred before Anthropic downloaded them and within 5 years of publication.

Turns out, authors are discovering that their books were not registered with the copyright office even though it was part of their contract that the publisher take care of that. This even applies to big publishers. Lesson here: If the publisher is supposed to register, check that they have. It’s easy to do. https://publicrecords.copyright.gov/

This is all of particular interest to me because turns out 3 of my books were used. I’ve done all the things required to be part of the settlement. We’ll see what happens. I’ll be sharing the money with my publisher since they are still publishing those books.

A new post on WriterBeware alerted me to a new twist. Apparently there are law firms encouraging authors to opt out of the settlement so they can be part of a different lawsuit and potentially get more money. ClaimsHero is one of these. It’s not guaranteed that an author would get more money or even that a new class action lawsuit would be certified by the court. 

Here are some posts you might find interesting:

Authors Guild post on lawsuit and what authors need to know: https://authorsguild.org/advocacy/artificial-intelligence/what-authors-need-to-know-about-the-anthropic-settlement/ 

Writers Beware post: https://writerbeware.blog/2025/10/31/the-anthropic-class-action-settlement-what-you-need-to-know-right-now/

AI is here to stay. As far as I’m concerned, it’s a tool. Like any other tool, it’s how you use it that is important. I do believe, though, that AI companies should get an author’s permission before using their works. 

What do you all think of this?