The Journal.
The Journal.

Sep 8, 2026 · 21 min

AI training turns obscure books into disposable data

AI Companies Are Buying and Destroying Used Books

The episode examines whether AI companies’ demand for training material is quietly removing rare books from public circulation.

3 key takeaways
  1. 1Used-book sellers trace mysterious bulk orders from obscure titles to scanning facilities and, ultimately, recycling.
  2. 2Court filings in Anthropic’s copyright case describe Project Panama, a mass physical-book acquisition and scanning effort.
  3. 3Booksellers must weigh unexpected income against the possibility that AI demand will permanently erase uncommon works from circulation.

Don't miss

A hidden tracker follows a purchased book from an Ohio warehouse through an Illinois scanning facility to a paper recycler in Mexico.

The brief

Used-book sellers begin receiving unusually large orders for obscure nonfiction, bringing welcome revenue but raising questions about buyers with unfamiliar names.

One bookseller hides tracking devices in the books and follows a shipment from an Ohio warehouse to an Illinois scanning facility, then to a paper recycler in Mexico.

Court filings in Anthropic’s copyright case reveal Project Panama, in which millions of physical books were bought and scanned for AI training, often through destructive methods.

The episode’s central tension is practical and cultural: sellers benefit financially, yet repeated AI purchases could remove rare books from circulation and weaken the used-book trade.

Melissa Korn and Jessica Mendoza consider why a physical used book carries a cultural and emotional value that extracted text cannot replace.

Books & mentions

Listen to the full episode and explore every guest, topic, and moment on PodLume.

What is PodLume?

PodLume turns podcasts into searchable knowledge. AI-decoded transcripts, identified guests and topics, smart highlights, and cross-show search across the world’s best conversations — all in your pocket.

AI training turns obscure books into disposable data | PodLume