Court Documents Reveal Anthropic's Secret Plan to 'Destructively Scan All the Books in the World'

Key points

• A federal court case in northern California, Bartz v Anthropic PBC, exposed an internal Anthropic project called "Project Panama" in late July 2025.
• An internal memo described the project's goal as "destructively scan all the books in the world" to gather training data for Claude, Anthropic's AI assistant.
• The same memo warned staff to use the codename because Anthropic did not "want it to be known" that the work was happening.
• The project surfaces as AI companies face growing legal pressure over whether using copyrighted books to train AI systems requires permission from authors.

Anthropody builds Claude, a large language model, meaning software trained on vast amounts of text that can then hold conversations, answer questions and write. Like all such systems, Claude needs enormous quantities of written material to learn from. Where that material comes from is now the subject of lawsuits across the United States.

Court exhibit 21 in Bartz v Anthropic, filed in the Northern District of California, contains an internal memo that asks, bluntly: "What is Project Panama?" The answer written inside it: "Project Panama is our effort to destructively scan all the books in the world."

"Destructively" here is a technical term from the book-scanning world. It means cutting the spine off a physical book so the pages can be fed through a scanner quickly. The book is ruined in the process. It is faster and cheaper than flatbed scanning, but it is still scanning a copyrighted object without the author's consent.

The memo does not stop at describing the project. It also explains why a codename was used. "[B]ecause we don't want it to be known that we are working on this," it reads. That line now sits in a public court filing.

Why does this matter to authors and readers?

Authors whose books were scanned get no say and no payment. Copyright law in the United States gives creators the exclusive right to reproduce their work. Whether feeding a scanned book into an AI training pipeline counts as that kind of reproduction is precisely what courts are working through right now.

The Bartz case, first reported by The Guardian, is one of several active suits testing this question. A ruling against Anthropic could force it to license books it has already used, pay damages, or change how it collects training data in future.

For readers and Claude users, the immediate practical effect is small: the service keeps running while litigation proceeds. But if courts consistently side with authors, the cost of training AI systems could rise sharply, which would affect which companies can afford to build them.

What happens next?

The case is still active. No final judgment on liability has been handed down. Anthropic has not commented publicly on the Project Panama memo.

Authors who believe their work was used without consent can check whether their titles appear in datasets that researchers have made public, though no complete list of Anthropic's training sources exists yet.

The broader legal fight over AI training data is moving fast. Congress has held hearings, and several bills are proposed, though none have passed, that would require AI companies to disclose what copyrighted material they trained on.

Common questions

Does Anthropic admitting this mean it broke the law?

Not automatically. Using copyrighted material for AI training may qualify as "fair use", a legal exception that allows limited reproduction without permission. Courts are still deciding whether that defence applies here.

Could this affect other AI companies?

Yes. Similar lawsuits target OpenAI, Meta and others over training data. A ruling in Bartz would set a precedent that all of them are watching closely.