Anthropic, an AI company backed by Amazon and Alphabet, recently agreed to a $1.5 billion settlement. This payout resolved a lawsuit over the use of pirated books to train its chatbot, Claude.
The lawsuit was filed in 2024 by Andrea Bartz, a novelist, along with two other authors. They claimed Anthropic used unauthorized copies of their works to teach Claude how to understand and generate text.
A federal judge approved the settlement on July 21, 2026. It is the largest known copyright payout in US history. Thousands of authors and publishers will receive payments averaging about $3,000 per book.
The settlement covers around 482,000 works. Authors have claimed about 91% of those so far. The payments will be spread out over the last half of the financial year, ensuring steady compensation.
What Anthropic Must Do Next
As part of the settlement, Anthropic must destroy all original and duplicate copies of pirated books used for training. These files came from known pirated book repositories.
Even after this, authors keep the right to sue if Anthropic uses their works without permission in the future. This sets a new precedent for AI companies using copyrighted material.
US District Judge Araceli Martinez-Olguin said the settlement gave “meaningful relief” to authors and publishers affected by the unauthorized use of their work. The ruling aims to protect creators while AI evolves.
Voices from the Case
Plaintiff attorney Justin Nelson expressed eagerness to make payments quickly. He said, “We look forward to making distributions to the Class as promptly as possible.”
Aparna Sridhar, speaking on behalf of the class, noted the high claim rate from authors and publishers. She said, “We are pleased that more than 91% of authors and publishers covered by the settlement have claimed their share of the payment.”
This settlement highlights a growing tension. AI companies want massive data sets to teach their models. But much of that data includes copyrighted content. Old books, once ignored, became valuable sources because their text is “clean” and rich.
Anthropic’s case signals more scrutiny over how AI firms collect training data. The industry now faces pressure to respect copyrights or face costly lawsuits.
For authors and publishers, the settlement offers a rare win. It shows their work has real value, even in the age of machines. And it warns AI developers to rethink how they gather data to build smarter, fairer tools.
Based on
- AI firms are buying up old books because they are the last slop-free data left — thenextweb.com
- Bloomsbury set for payout from £1.1 billion Anthropic US copyright settlement | The Independent — independent.co.uk
- Anthropic to pay authors $1.5 billion in largest-ever copyright settlement | Fortune — fortune.com
- Judge approves Anthropic’s $1.5B settlement for authors after using pirated books to train AI | The Independent — independent.co.uk
- Anthropic to pay US$1.5B over pirated books used to train Claude — ctvnews.ca
