Anthropic agrees to pay $1.5 billion to settle authors copyright suit

In short: Anthropic agreed to pay $1.5 billion to settle a lawsuit from authors who said it used their books without permission to train its AI.

What happened

Anthropic, an AI company, agreed to pay $1.5 billion to settle a class-action lawsuit brought by authors. The authors accused the company of copying more than 7 million digitized books and using them to train its AI models.

Training an AI model means feeding it a huge amount of text so it can learn patterns and produce new writing later. You can think of it like teaching a student by giving them a giant library to read, except the lawsuit claims many of these books were taken without permission.

The settlement could compensate about half a million authors. Reports say it works out to around $3,000 per work used. Anthropic did not admit wrongdoing or legal liability as part of the agreement.

One of the lead plaintiffs, author Charles Graeber, said Anthropic used his books The Good Nurse and The Breakthrough without permission in a database used for training.

Why it matters

This case is a sign that the rules around AI and copyrighted writing are starting to get clearer. If courts and settlements keep saying companies must pay for unlicensed books, it could change how AI systems are built and what they are allowed to learn from. Similar copyright lawsuits are still ongoing against several AI companies, including Anthropic, OpenAI, Google, Meta, xAI, and Perplexity AI.

Source: NYTimes