A federal judge has granted final approval to Anthropic’s $1.5 billion settlement with authors and publishers who accused the artificial intelligence company of unlawfully obtaining copyrighted books.
U.S. District Judge Araceli Martinez-Olguin approved the agreement on July 20, 2026, bringing one of the most closely watched copyright disputes involving generative artificial intelligence to a conclusion.
The settlement is the largest known resolution of a copyright case in the United States. It covers hundreds of thousands of books and is expected to provide approximately $3,000 for each eligible work, with payments shared among the authors and publishers who hold the relevant rights.
The class action lawsuit challenged the way Anthropic assembled a large collection of books for developing its artificial intelligence systems, including Claude.
The company obtained books through different methods. Some physical books were legally purchased, scanned and converted into digital files. Other books were downloaded from unauthorized online libraries, including Library Genesis and Pirate Library Mirror.
The distinction between these sources became central to the court’s decision.
Judge William Alsup, who handled the case before retiring, ruled that using copyrighted books to train artificial intelligence models could qualify as fair use under the circumstances considered in the case. However, he found that downloading and keeping pirated copies of books created a separate copyright problem.
Claims connected to the unauthorized acquisition and storage of those books were allowed to proceed toward trial.
Anthropic agreed to the $1.5 billion settlement in 2025 before the piracy-related claims reached a jury.
Going to trial could have exposed the company to significantly larger damages, depending on the number of infringed works and the jury’s findings. The settlement allowed Anthropic to resolve the claims without facing that uncertainty.
The agreement received preliminary approval from Judge Alsup in September 2025. Judge Martinez-Olguin later reviewed the proposed payments, legal fees and objections before granting final approval.
More than 90 percent of the works covered by the agreement have reportedly been claimed by eligible authors and publishers.
The case produced an important distinction for the artificial intelligence industry.
The court treated the use of lawfully obtained books for AI training differently from the act of acquiring books through piracy. This means the settlement should not be interpreted as a broad ruling that all training of AI systems on copyrighted material is illegal.
At the same time, the outcome shows that companies may face substantial liability when copyrighted training material is collected through unauthorized sources.
The fair use ruling was issued by a federal district court and does not create a binding nationwide precedent. Other courts remain free to reach different conclusions based on the facts of the cases before them.
Because Anthropic settled the remaining claims, the dispute will not proceed to an appeals court that could have produced a broader and more authoritative ruling.
Final approval closes this particular lawsuit, but it does not resolve the wider conflict between copyright owners and artificial intelligence companies.
Authors, publishers, artists and media organizations have filed numerous lawsuits challenging the use of protected material to develop generative AI systems. Companies including Google, Meta, Midjourney and OpenAI are facing separate claims involving copyrighted training data.
Those cases could lead to different interpretations of fair use, licensing requirements and responsibility for obtaining training material legally.
For authors and publishers included in the Anthropic settlement, the court’s approval clears the way for compensation. For the AI industry, the case delivers a clear warning that even when model training may be protected as fair use, the methods used to collect training data can still carry major legal and financial consequences.
Share your thoughts about this article.
Be the first to post a comment!
