Judge Slashes Lawyer Fees to $101M in $1.5B Author-AI Copyright Deal

The gavel finally came down on the publishing industry’s biggest AI showdown. A federal judge approved a $1.5 billion settlement between Anthropic and a class of authors on Monday, officially closing the largest copyright class-action ever certified. The cash figure is historic, but the real story is in the cuts the judge made on her way to signing off.

US District Judge Araceli Martínez-Olguín didn’t just rubber-stamp the deal. While she rejected a last-ditch effort by a handful of authors to blow up the agreement—calling their objections meritless—she also took a cleaver to the legal fees. The plaintiffs’ lawyers initially angled for 20 percent of the pot, or $300 million. They later dropped that ask to 12.5 percent. Martínez-Olguín went much further, capping the payout at less than 7 percent, roughly $101 million. It’s a clear signal that in mega-class actions, the bench is watching the spread between what victims get and what their attorneys take home.

The participation numbers made the authors’ objections a tough sell. The judge pointed out that 95 percent of the class received notice, and a striking 91 percent of impacted authors and publishers had already filed claims. Only 350 people opted out entirely, while just 54 tried to object or filed late requests to leave the class. When you have that level of buy-in, convincing a judge the deal is unfair becomes a steep climb.

The per-work payout lands at an estimated $3,000—a number Martínez-Olguín noted is four times the minimum statutory damages. That context matters, especially since the court had previously ruled that training AI on the books might be fair use, but the initial piracy of the texts probably wasn’t. The settlement avoids a messy trial over that very distinction. For the vast majority of writers who didn’t want to gamble on a novel legal theory, a quick, guaranteed four-figure check per book was apparently good enough.