Anthropic's $1.5B copyright settlement wins final approval
A federal judge has given final approval to Anthropic's $1.5 billion class-action copyright settlement with authors and publishers, letting payouts begin. The anthropic8217s landmark 15b copyright deal pays about $3,000 per work across roughly 500,000 titles, though it does not settle AI-training legality industry-wide.
Key Takeaways
- Judge Araceli Martinez-Olguin granted final approval Monday to Anthropic's $1.5 billion class-action settlement with authors and publishers.
- Eligible rights holders are expected to receive about $3,000 per work across an estimated 500,000 works.
- A prior ruling treated AI training on copyrighted text as fair use, but found Anthropic's use of pirated books illegal.
- The case ends without appellate precedent, leaving AI copyright fights against Google, Meta, Midjourney, and OpenAI unresolved.
Anthropic can finally start cutting checks after a federal judge signed off on the landmark deal, according to TechCrunch's report citing Reuters. The settlement is believed to be the largest in U.S. copyright history and is a major marker for the wider Future Tech & AI Wonders beat.
What did the court approve, and who gets paid?
Judge William Alsup of the U.S. District Court for the Northern District of California gave the settlement preliminary approval last year. He had ruled that Anthropic illegally downloaded and stored millions of copyrighted books.
Alsup has since retired. Judge Araceli Martinez-Olguin issued the final approval on Monday, clearing the way for payments to authors and publishers who hold rights in the covered works.
The fund is structured to deliver roughly $3,000 per work across an estimated 500,000 works. Those amounts are shared among the rights holders tied to each title.
Why don't many authors see this as a win?
The size of the check is historic, but the legal reasoning still favors Anthropic on the core training question. Alsup ruled that training an AI model on copyrighted text counts as fair use, a result widely viewed as a turning point for AI labs.
What the court did not excuse was how Anthropic built part of its library. Books the company purchased and scanned were treated as fine. Books downloaded from pirate sites such as Library Genesis and Pirate Library Mirror were not.
Alsup found the second method illegal on its own terms and said that piracy question could go to trial. Anthropic settled soon after to avoid a jury and the damages that might have followed.
Does this settle AI copyright law for everyone?
No. Final approval closes this case, but Alsup's fair-use finding was a single district-court decision. Because Anthropic settled, the dispute will not reach an appeals court that could create binding precedent.
Other judges remain free to decide their own cases on their own facts. Copyright suits against Google, Meta, Midjourney, and OpenAI are still pending over whether training on copyrighted works is lawful.
Just last week, publishers and authors including Hachette, Cengage, Elsevier, Scott Turow, and S.C.R.I.B.E. filed a class action accusing Google of using their works to train Gemini. That fight underscores why Monday's Anthropic approval is a landmark payment, not a final industry rulebook.