---
title: "Amazon Destroys Rare Books to Train AI Models"
source: "TechCrunch：AI（RSS）"
source_tier: "t2"
url: "https://techcrunch.com/2026/08/17/amazon-once-an-online-bookseller-is-destroying-rare-books-to-train-ai-models/"
permalink: "https://feed.astralor.com/items/itm_966ff312c3d31c897dfd6cac"
published: "2026-08-17T16:38:44.000Z"
score: 68
channel: "ai"
tags: ["AI training data","Amazon","Large Language Models","Cultural heritage preservation"]
---

# Amazon Destroys Rare Books to Train AI Models

## AI commentary

> This news is worth watching because AI giants are now reaching into precious physical documents, reflecting the predicament of shrinking public data pools. For AI practitioners, it signals that the future data race may rely more on non-public or copyrighted content, and urges us to discuss fair use boundaries early.

## At a glance

According to TechCrunch, Amazon is destroying rare books to obtain high-quality training data for AI models. As public text on the internet has been nearly exhausted by large models, these rare volumes—carrying unique historical and scholarly value—have become a critical resource. However, the practice also triggers intense discussion over cultural heritage preservation, copyright boundaries, and the information monopoly of tech giants. Publishers and academics have expressed strong concerns, demanding more responsible approaches to data acquisition and greater transparency from companies like Amazon. This AI training approach at the expense of cultural heritage underscores the desperation of tech companies facing data scarcity. There has been no official response yet.

## Key points

- Amazon is converting rare books into AI training data, despite their significant cultural and historical value.
- Rare books provide unique high-quality content for LLMs, filling gaps left by public online data.
- The practice has sparked controversy among publishers, academics, and the public over cultural heritage preservation and fair use.

## Full text

Rare books are incredibly valuable for training LLMs, since these models have already trained on whatever's available online.

---

Verify against the original before citing; curated by Astral Feed: https://feed.astralor.com/items/itm_966ff312c3d31c897dfd6cac