History is being made in the world of AI and copyright, and it’s all about to hit authors’ bank accounts in a big way. Anthropic, one of the major names in artificial intelligence (you know, the team behind the popular AI chatbot Claude), just agreed to pay at least $1.5 billion (yes, you read that correctly, ‘billion’) to authors in a massive class-action settlement over how their work was used to train AI models.
What’s truly incredible is that this deal is likely the largest ever reported in a US copyright case. The settlement (pending court approval), works out to roughly $3,000 per book or work that was used. If more than 500,000 works end up covered, Anthropic will owe even more. The settlement also requires Anthropic to destroy all the original and copied files it downloaded, and more importantly, no reusing those texts for future AI training.
What’s not included? Anthropic isn’t getting a free pass moving forward. The settlement only covers past use, which means anything created or used after late August 2025 is not immune from new claims or lawsuits.
This drama’s been a year in the making. Authors first sued in 2024, accusing Anthropic of building its AI by “stealing” hundreds of thousands of copyrighted books. After some legal back-and-forth, the court ruled that Anthropic didn’t break the law by training AI on legally purchased books, but did have to answer for using pirated books—which is where this class-action settlement comes in.
So what happens next? Any author or rightsholder who thinks their work was involved can find out more (and potentially cash in) at the AnthropicCopyrightSettlement.com website. And as AI companies face more and more copyright lawsuits, this case is likely just the first of many big payouts to come.
In a nutshell: if you write for a living, AI firms are starting to pay for using your words—and the numbers are only getting bigger.
