Connect with us

AI

Judge lets Anthropic train AI on books without author consent

The ruling allows companies to train AI models using purchased books under fair use laws.

Smartphone scanning books with Google Lens app
Image: Google

Just a heads up, if you buy something through our links, we may get a small share of the sale. It’s one of the ways we keep the lights on here. Click here for more.

A federal judge, William Alsup, recently ruled that the AI company Anthropic did not break the law by using published books to train its artificial intelligence systems without the authors’ permission. 

This is the first time a court has officially supported the idea that “fair use” laws can protect AI companies when they use copyrighted materials to teach their models how to understand and generate text.

This ruling is a setback for authors, artists, and publishers, many of whom are suing major tech companies like OpenAI, Meta, and Google for similar reasons. 

Although this decision only applies to one case and doesn’t set a nationwide rule, it gives other judges a possible reason to side with tech companies in future cases.

At the heart of these lawsuits is the concept of “fair use,” a legal rule that allows people to use copyrighted material without permission in certain cases, like for parody, education, or research.

 However, the law hasn’t been updated since 1976, long before the internet or AI tools like ChatGPT existed. 

Whether something counts as fair use often depends on how it’s used, if it’s for profit, and how different the new work is from the original.

In this specific case, called Bartz v. Anthropic, the authors also accused Anthropic of using illegal methods to get their books. 

The lawsuit claims the company downloaded millions of books from pirate websites and stored them to build a permanent library of the world’s written works. 

While Judge Alsup said using the books for training AI could count as fair use, he made it clear that the issue of illegally downloading those books isn’t settled. (Via: Tech Crunch)

The court will now move forward with a trial to decide whether Anthropic broke the law by building this “central library” using pirated books and what financial penalties they might face. 

As Judge Alsup stated, buying the books later doesn’t erase the damage done by downloading them illegally in the first place.

What do you think about this ruling? Do you think copyright rules shouldn’t apply to AI training? Tell us what you think below in the comments, or via our Twitter or Facebook.

Follow us on Flipboard, Google News, or Apple News

Ronil is a Computer Engineer by education and a consumer technology writer by choice. Over the course of his professional career, his work has appeared in reputable publications like MakeUseOf, TechJunkie, GreenBot, and many more. When not working, you’ll find him at the gym breaking a new PR.

Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

More in AI