Anthropic Settles AI Copyright Lawsuit for Record $1.5 Billion

Anthropic, which had been embroiled in a class-action lawsuit over allegations of unauthorized use of copyrighted works to train its artificial intelligence (AI) models, has settled the case for approximately 2 trillion won, the largest copyright settlement in U.S. history.

Judge Araceli Martínez-Olguín of the U.S. District Court for the Northern District of California issued an order on July 20 (local time) granting final approval to the settlement agreement submitted by both parties in the copyright infringement class action filed by authors against Anthropic.

Under the terms of the settlement, Anthropic will deposit $1.5 billion (approximately 2.2 trillion won) to be paid to the authors.

Members of the class action will receive approximately $3,000 (about 4.4 million won) in compensation per copyrighted work.

This amount is four times the $750 statutory minimum damages typically available in standard copyright infringement cases.

The court also ordered Anthropic to destroy any copyrighted materials or copies downloaded from so-called pirate websites.

Some plaintiffs participating in the class action raised objections, arguing that the compensation was too low, but their objections were dismissed.

The court explained that if the case had proceeded to a full trial instead of a settlement, there was no guarantee of a victory for the plaintiffs, and conversely, even if a large damages verdict were reached, procedural issues could arise.


Authors, including Andrea Bartz, author of the bestselling thriller novel "We Were Never Here," filed the class-action lawsuit in 2024, alleging that Anthropic used their works without permission during the training process for its AI model, "Claude."

However, the court previously ruled in favor of Anthropic, stating that using legally purchased books for AI training constitutes "fair use."

Nevertheless, the court determined that Anthropic's act of downloading books from pirate websites could constitute copyright infringement and proceeded with a damages trial specifically for those instances.

In response, Anthropic reached an agreement with the plaintiffs to pay a total of $1.5 billion, or approximately $3,000 per book, for the 500,000 books in question.

At the time, Anthropic admitted in a court filing that it chose to settle "under the pressure that this litigation could lead to a trial that might end our business."

Justin Nelson, an attorney representing the plaintiffs, stated in a release to Reuters, "We are pleased with the court's decision to approve this historic settlement," adding that "this is the largest copyright damages case in history."

Aparna Sridhar, Deputy General Counsel at Anthropic, also emphasized in a statement, "We hope to resolve this matter quickly," and added, "The ruling that AI training using books constitutes fair use under copyright law remains valid."

However, some authors and publishers remain outside of this settlement and are pursuing separate legal actions.

By concluding this copyright lawsuit with a settlement, Anthropic has removed a legal hurdle ahead of its planned initial public offering (IPO), which is targeted for the fourth quarter.

(Photo: Screenshot of evidence submitted to the U.S. District Court for the Northern District of California (Case No. 3:24-cv-05417), Yonhap News)
※ Please note: This article was translated by AI and may contain errors.