Anthropic’s Claude chatbot|@AnthropicAI|X
A trio of authors who sued the Claude chatbot maker Anthropic, alleging it trained the AI tool with pirated copies of books, have reached a $1.5 billion settlement.
The lawsuit was filed last year by authors Andrea Bartz, Charles Graeber, and Kirk Wallace Johnson, later representing a broader class of writers and publishers.
The settlement still needs a judge’s approval. The deal will compensate writers and publishers with about $3,000 each for an estimated 500,000 books—a total that lawyers say represents the largest copyright recovery to date.
In June, US District Judge William Alsup ruled that while training AI on copyrighted books wasn’t illegal, Anthropic had wrongfully obtained the material from pirate websites.
Evidence showed that the AI startup downloaded over 7 million books from online libraries, including Books3, Library Genesis (LibGen), and the Pirate Library Mirror. If Anthropic had gone to trial and lost, it could have been asked to shell out ~$1 trillion in damages.
Anthropic expects $5 billion in sales this year and has recently valued itself at $183 billion.