Anthropic settles AI copyright case brought by authors

Anthropic settles AI copyright case brought by authors

Reuters reported on August 27, 2025 that Anthropic settled the class action with a group of U.S. authors this week and that this was a first.

We recently added a clause to our Imprint pages stating that use of our authors’ material from any of our books, by language learning AI systems or training systems, must be obtained with permission from publisher and author/s. It’s the minimum one should expect from all current copyright laws including in Australia. These laws are in place to protect, not devalue, the work of human endeavour, creative ideas and original material.

This doesn’t mean publishers or content owners are against using AI. There are many applications of AI which will be important in the future: from ground breaking medical cures to the exciting creation of more original material such as Brian Eno's biopic. By all means, ask it to tell you how to analyse a paragraph, find a source, interrogate a phrase or find and make a list of incorrectly used semi-colons in a thesis of 50,000 words. 

A respect for copyright and the law is about preventing the theft of the work originated by others, and protecting those owners and publishers from lost attribution and income. We don’t allow it for artworks like paintings or music, we shouldn’t allow it for the written word. In addition, a change to our copyright laws to allow such theft would open up companies, as well as creators, to having their own IP stolen, copied and used by others freely without attribution. The basis of copyright in business also protects developments by inventors, founders and trade secrets.

Perhaps the Anthropic team have realised this. Write to your local minister if you’re an author, business or creator!

At a time when Australian copyright laws are under pressure and governments feel forced to review them under the guise of “productivity” gains, the Anthropic settlement is a significant development.

According to the article, legal experts said the case's distinct qualities complicate the deal's potential influence on a wave of ongoing copyright lawsuits against other artificial-intelligence focused companies like OpenAI, Microsoft, and Meta.

Backed by Amazon and Google, a trial was looming in December after a judge found Anthropic liable for pirating millions of copyrighted books.

The settlement, which requires a judge's approval, is not yet public. U.S. courts have just begun to grapple with copyright questions related to generative AI, which could prompt other defendants to hold out for favourable rulings.

Leave a comment

Comments will be reviewed before they are published.