Skip to content
Tools · Aug 17, 2026

Amazon buys rare books, destroys spines to scan for AI training

Amazon confirmed purchasing rare texts through commercial channels to improve AI products, per 404 Media investigation involving a tracking device sent to an Amazon facility in Las Vegas.

Trust72
HypeSome hype

1 source · cross-referenced

ShareXLinkedInEmail
TL;DR
  • Amazon confirmed to 404 Media it purchases books via commercial channels to improve AI products and services.
  • An investigation by 404 Media found Amazon cutting spines off rare books and scanning them at a Las Vegas facility for AI training data.
  • Rare books are valuable for AI training because they are often unavailable online and predate 2022, avoiding contamination with AI-generated text.

An investigation by 404 Media reported that Amazon is purchasing rare books, removing their spines, and scanning them at a facility in Las Vegas identified as VGT3, which uses a logo featuring a dinosaur holding a book. The report included documentation from a tracking device placed in a rare book that was delivered to the facility.

Amazon confirmed to 404 Media that it "purchases books through commercial channels to improve the products and services customers use," framing the activity as part of efforts to enhance AI systems.

The report noted that rare books are particularly valuable for training large language models because many are out of print or unavailable online, and their publication dates predate 2022, reducing the risk of contamination with AI-generated text that could degrade model quality.

The practice underscores the broader industry challenge of securing large volumes of high-quality training data as models exhaust readily available online sources and seek alternative, often physical, text sources.

Sources
  1. 01TechCrunch — AIAmazon, which started off selling books, is destroying rare texts to train AI
Also on Tools

Stories may contain errors. Dispatch is assembled with AI assistance and curated by human editors; despite the trust-score filter, mistakes happen. We correct publicly — every article links to its revision history. Nothing here is financial, legal, or medical advice. Verify before relying on any claim.

© 2026 Dispatch. No ads. No sponsorships. No paid placement. Reader-supported via Ko-fi.

Built by a person who cares about honest AI news.