Anthropic agrees to pay $1.5 billion to authors over books used for AI training

Anthropic has reached a landmark settlement, agreeing to pay $1.5 billion to authors whose books were downloaded and used without permission to train its AI chatbot Claude. That equates to about $3,000 per work for around 500,000 identified books, with potential increases if more are found. Attorneys for the authors described this as the largest copyright recovery in history and a landmark case in the AI era. The company has committed to destroying the downloaded copies from unauthorized sources but does not admit liability.

📎 Motion for Preliminary Approval (PDF)