Anthropic ordered to pay $1.5 billion in AI copyright settlement

A legal ruling involving Anthropic has addressed the complexities of copyright in the era of artificial intelligence. Judge William Alsup ordered the technology company to pay a $1.5 billion settlement to a group of writers whose works were used to train AI models.

Despite the significant settlement, the judge ruled that Anthropic’s AI training process was legal. The penalty was specifically issued because the company obtained books from illegal online “shadow libraries.” The ruling compared the way large language models (LLMs) process data to a human writer studying literature to create something new, rather than simply copying or replacing the original works.

Legal expert Cathy Gellis noted that the decision may ultimately favor AI companies. She suggested that the $1.5 billion fine is relatively minor compared to the projected annual revenues of major AI firms, and that the court’s view of AI training as being akin to reading copyrighted works provides a favorable precedent for the industry.