Anthropic may finally start cutting checks to a group of authors and book publishers who are suing the AI lab for copyright infringement. A federal judge on Monday gave final approval to Anthropic’s $1.5 billion class-action copyright settlement. This was reported by Reuters.
Judge William Alsup of the US District Court for the Northern District of California granted preliminary approval of the settlement after ruling last year that Anthropic illegally downloaded and stored millions of copyrighted books.
Alsup has since retired, and Judge Araceli Martinez-Olguin signed off on the settlement on Monday.
The fee will be $3,000 per work on an estimated 500,000 works and will be shared between authors and the publishers who hold the rights to them. It is believed that the settlement The largest in the history of US copyright law, many authors and creators still do not see this as a victory.
It depends on how the legal issue is resolved. Alsup sided with Anthropic on the main issue. It ruled that training an AI model on copyrighted text would be considered fair use—a decision that was seen as a turning point for the AI industry. But the ruling did not excuse how Anthropic acquired the books in the first place. Anthropic built its training library from two sources: (good) books it bought and scanned, and books it downloaded from pirate sites like Library Genesis and Pirate Library Mirror. Alsup considered the second method illegal on its own terms, and said that the issue of piracy could be litigated; Anthropic quickly agreed to a settlement to avoid a trial and any damages a jury might award.
Although the final approval closes the case, Alsup’s decision does not resolve the legal question because it is a single district court decision, and Anthropic’s decision to settle means the case will never reach an appellate court to become binding precedent.
Other judges are still free to make their own findings of fact, which is what happens elsewhere. There are still a number of copyright claims against companies such as Google, Meta, Midjourney and OpenAI over whether it is legal to train AI models on copyrighted works. Last week, a group of publishers and authors including Hachette, Cengage, Elsevier, author Scott Turow, and SCRIBE class action lawsuit Google is accused of using the company’s copyrighted works to train its artificial intelligence platform Gemini.
When you purchase through links in our articles, we may earn a small commission. This does not affect our editorial independence.
