The landmark piracy case between authors and Anthropic over the company’s training of its Claude AI chatbot is nearing its conclusion, with almost 100,000 claims filed.
Plaintiffs in Bartz v. Anthropic filed a motion for final approval for the settlement agreement, stating, «The class has spoken.» Anthropic is expected to pay $1.5 billion in total, and authors will receive a payout of $3,000 per qualifying work. The settlement was approved in September 2025.
The lawsuit, which initially focused on copyright infringement, alleges that Anthropic illegally downloaded over 500,000 copyrighted works from pirated content websites Library Genesis and Pirate Library Mirror to train Claude.
Since the process opened, nearly 100,000 claims have been filed, and less than 0.5% of the class members have opted out of the settlement.
Under the settlement agreement, Anthropic has certified that it did not use any pirated content in any released Claude model and has promised to destroy all pirated copies of class members’ works.
US District Judge William Alsup initially withheld the approval of the settlement after being concerned that qualifying authors in the lawsuit wouldn’t have time to join. This resulted in an extensive outreach program to authors who qualified for payment.
The campaign included traditional mail, email campaigns, social media and other digital outlets, as well as professional and informal author and publisher networks. The official settlement website also includes a searchable database of qualifying works and full details, including important documents and dates.
The court is scheduled to hear the motion for final approval on April 23, and the last day for class members to file a claim is March 30.
The final motion includes a series of statements in support of the agreement from writers’ groups and organizations in what will become the largest copyright payout in history. The case is the first of its kind with a major artificial intelligence company to reach a settlement.
