Anthropic Pays $1.5B for Pirated Books Settlement
- The AI company Anthropic has agreed to a settlement exceeding $1.5 billion to resolve a class action lawsuit alleging the use of copyrighted books to train its Claude...
- Anthropic, a leading AI company developing the Claude chatbot, faced a class-action lawsuit filed by authors Andrea Bartz, Charles Graeber, and Kirk Wallace johnson.
- Plaintiffs argued that Anthropic's actions infringed on their copyrights and devalued their work.
Anthropic Settles Class Action Lawsuit Over AI Training data for $1.5 Billion
Table of Contents
The AI company Anthropic has agreed to a settlement exceeding $1.5 billion to resolve a class action lawsuit alleging the use of copyrighted books to train its Claude AI models. The case, brought by authors, sparked debate over fair use in the age of artificial intelligence.
Published September 6, 2025, at 22:12:37
the Lawsuit and the Allegations
Anthropic, a leading AI company developing the Claude chatbot, faced a class-action lawsuit filed by authors Andrea Bartz, Charles Graeber, and Kirk Wallace johnson. The authors accused Anthropic of illegally copying thier books to train Claude, a direct competitor to OpenAI’s ChatGPT [OpenAI website].The core of the complaint centered on whether using copyrighted material to train AI constitutes fair use.
Plaintiffs argued that Anthropic’s actions infringed on their copyrights and devalued their work. The lawsuit alleged that Anthropic not only used legally purchased books but also downloaded millions of pirated copies to build a comprehensive training dataset.
The Court’s Ruling on Fair Use
In June, U.S. District Court judge William Alsup issued a partial ruling in favor of Anthropic. Judge Alsup steadfast that the process of training AI models with books – regardless of whether those books were purchased legally or obtained through piracy – was “transformative” enough to qualify as “fair use” under copyright law [CourtListener Docket].
The judge likened AI training to the way humans learn by reading, arguing that the AI models were not simply reproducing the original works but creating something new. He stated, “The technology at issue was among the most transformative many of us will see in our lifetimes.”
Though, Judge Alsup rejected Anthropic’s attempt to secure blanket protection for its practices. He specifically ruled that the company’s practice of downloading millions of pirated books to create a permanent digital library was *not* justified by fair use. This distinction proved crucial in the subsequent settlement negotiations.
The $1.5 Billion Settlement
The $1.5 billion settlement, announced on September 6, 2025, represents a meaningful outcome in the ongoing debate surrounding copyright and AI. Plaintiffs’ attorney Justin nelson hailed the agreement as “a landmark settlement far surpasses[ing] any other known copyright recovery” and “the first of its kind in the AI era” [Reuters Report].
The settlement details are still subject to court approval, but it is expected to establish a framework for compensating authors whose works were used in the training of Anthropic’s AI models. The exact distribution of funds among the class members will be determined