Anthropic Reaches Settlement in Landmark AI Copyright Case with US Authors

A smartphone screen displays the introduction to Claude, an AI assistant by Anthropic, with options to submit business interest or talk to Claude. The word "Anthropic" is blurred in the background.

Artificial intelligence company Anthropic has reached a settlement in a high-profile class action lawsuit brought by a group of U.S. authors who alleged the company used pirated books to train its AI systems without permission or compensation.

The agreement, disclosed in a court filing on Tuesday, avoids a potentially massive damages trial that had been scheduled for December and has a potential knock-on effect on the debate around whether human creatives should be compensated when their work is used to train AI.

In June, U.S. District Judge William Alsup ruled that while Anthropic’s use of the materials to train its AI assistant Claude qualified as “fair use” because it was “quintessentially transformative,” the company violated authors’ rights by creating a central library of pirated books. The judge ordered that the question of damages for this acquisition practice proceed to trial.

The lawsuit, filed last year by writers Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, accused Anthropic of unlawfully downloading approximately seven million books from so-called “shadow libraries,” including sites like LibGen. The plaintiffs argued that this constituted large-scale copyright infringement, with statutory damages under U.S. law potentially reaching billions of dollars.

On Tuesday, both parties informed the court that they had negotiated a proposed class settlement, with requests for preliminary approval due by early September.

“This historic settlement will benefit all class members,” says Justin Nelson, an attorney for the authors. He added that details of the settlement would be announced “in the coming weeks.” Anthropic has declined to comment.

Legal experts tells Reuters the resolution could influence the growing number of copyright lawsuits involving AI companies.

“The devil is in the details of the settlement and future litigation about the terms of the settlement,” says Shubha Ghosh, a professor at Syracuse University College of Law.

The Authors Guild recently began notifying writers who may qualify to participate in the settlement, with a list of affected works expected to be submitted to the court this week.

Anthropic, which has faced other copyright-related legal challenges — including a separate case brought by major record labels alleging illegal use of song lyric — did not disclose any financial terms of the agreement.

Settlements of this kind do not set legal precedent but may shape the course of ongoing litigation in the fast-evolving field of AI and copyright.


Image credits: Header photo licensed via Depositphotos.

Discussion