In a groundbreaking legal settlement, Anthropic has agreed to pay $1.5 billion to authors whose books were used without permission to train its AI chatbot Claude.

The landmark agreement could signal a turning point for creative professionals battling against tech companies using their work to develop artificial intelligence. Under the proposed settlement, authors will receive approximately $3,000 for each of an estimated 500,000 books used in training the AI model.

Led by thriller novelist Andrea Bartz and supported by other writers, the lawsuit exposed how Anthropic downloaded millions of digitized books from pirate websites like Library Genesis and Pirate Library Mirror. Federal Judge William Alsup’s previous ruling found that while AI training might qualify as “fair use,” the company’s method of acquisition was problematic.

Mary Rasenberger, CEO of the Authors Guild, celebrated the settlement as “an excellent result” that sends a strong message to the tech industry about respecting intellectual property. The settlement requires Anthropic to destroy the original book files and compensate authors for their unauthorized use.

This legal action highlights the ongoing tension between technological innovation and creators’ rights in the rapidly evolving AI landscape. Despite being valued at $183 billion and expecting $5 billion in sales this year, Anthropic found itself confronting significant legal and ethical challenges around content acquisition.

The settlement could potentially influence similar ongoing lawsuits against other AI companies like OpenAI, Meta, and Midjourney, suggesting a potential shift in how tech companies approach intellectual property in AI development.

As AI continues to transform various industries, this case underscores the critical importance of ethical data sourcing and respecting creators’ rights in the digital age.