cd /news/large-language-models/ai-firms-are-buying-up-old-books-the… · home topics large-language-models article
[ARTICLE · art-81747] src=novaramedia.com ↗ pub= topic=large-language-models verified=true sentiment=↓ negative

AI Firms Are Buying Up Old Books, Then Scanning and Destroying Them

Anthropic, the company behind Claude AI, has spent tens of millions of dollars acquiring millions of antiquarian books to scan for training its large language models, then destroying them, according to a Washington Post report. Internal documents unsealed in January reveal the project, codenamed Project Panama, used a hydraulic cutting machine to slice book spines and pages for industrial scanning, with Anthropic acknowledging it did not want the effort to be known. A judge recently approved a $1.5bn (£1.1bn) payment in a class action lawsuit brought by authors whose pirated work was used to train Claude, though a prior ruling found the use of physical books was protected as fair use under US law.

read2 min views3 publishedJul 31, 2026
AI Firms Are Buying Up Old Books, Then Scanning and Destroying Them
Image: source

‘We don’t want it known that we’re doing this.’ #

by Tom Midlane

29 July 2026

The world’s biggest AI companies are buying antiquarian books en masse so they can be scanned to train their large language models (LLMs) before being destroyed.

Anthropic, the company behind Claude AI, has spent tens of millions of dollars acquiring millions of books to feed its demand for human-authored training material, the Washington Post reported earlier this year.

Under the codename Project Panama, the firm used a hydraulic powered cutting machine to lop off books’ spines and then slice individual pages, which could then be scanned with industrial-grade imaging equipment before being discarded.

Internal documents unsealed in legal filings in January revealed that Anthropic was acutely aware of how negatively such a project would be received outside the company.

“Project Panama is our effort to destructively scan all the books in the world,” an internal planning document said. “We don’t want it to be known that we are working on this.”

Earlier this month a judge approved a $1.5bn (£1.1bn) payment in a class action lawsuit brought against Anthropic by a group of authors whose pirated work had been used to train Claude.

However, in the original ruling last year, a judge found that Anthropic’s use of physical books was protected as fair use under US law, Futurism reported.

The company was able to evade copyright issues by hiding behind a legal rule known as first-sale doctrine, which allows a buyer to do what they want with an item after purchasing.

The doctrine enables the second-hand book trade to exist – and a new market has now sprung up explicitly catering to AI companies seeking to buy vast quantities of old books.

“I personally have mixed feelings about all of this,” one bookseller told 404 Media. “It benefits me financially as well as by clearing out old inventory that is otherwise unlikely to sell.

“I’ve been well-suited for these sales with inventory from overseas and foreign language books. On the other hand, I don’t like the end-use, and I don’t like that uncommon books are being pulped.”

Tom Midlane is a freelance journalist.

── more in #large-language-models 4 stories · sorted by recency
── more on @anthropic 3 stories trending now
sponsored brought to you by zahid.host 4,200+ EU-deployed projects
reading about agents? ship yours in a single git push.

Run your AI side-project on zahid.host

EU-based hosting, git-push deploys, automatic HTTPS, no cold starts. Free tier with a custom domain — perfect for shipping the agent you just read about.

$git push zahid main
Live at https://your-agent.zahid.host
Get free account → Pricing
from €0/mo · no card required
LIVE [news/ai-firms-are-buying-…] indexed:0 read:2min 2026-07-31 ·