Anthropic has settled a lawsuit that accused the company of using pirated books to train its artificial intelligence models. The settlement amount, however, remains unclear: reports in the article title put the figure at $2 billion, while the article text itself states $1.5 billion.
The allegations
The lawsuit, filed by a group of authors and publishers, claimed that Anthropic copied copyrighted books from pirate websites to build its training datasets. The plaintiffs argued that the company had no license to use the works and that the practice violated copyright law. Anthropic did not admit wrongdoing as part of the settlement, according to the terms described in the reports.
Conflicting settlement figures
The discrepancy in the reported settlement numbers has not been explained. The headline of the article that broke the news cites $2 billion, but the body of the same piece says $1.5 billion. Neither Anthropic nor the plaintiffs’ legal team has publicly confirmed the exact amount. The confusion leaves the financial impact of the case uncertain.
What the settlement means for AI training
The case is one of several legal challenges facing AI companies over the use of copyrighted material for training. The settlement could set a precedent for how much companies are willing to pay to resolve such disputes out of court. It also highlights the ongoing tension between content creators and AI developers over data sourcing practices.
Anthropic, which is valued at $1.25 trillion in a December prediction market with a 91.5% YES probability, has not commented on the settlement beyond the terms included in the court filing. The company’s valuation prediction, however, is unrelated to the lawsuit and based on separate market speculation.
The next step is a court hearing scheduled to approve the settlement terms. Until then, the exact amount Anthropic will pay remains an open question.
