AI Book Burning? Companies Are Destroying Millions of Books to Feed Chatbots

What happened

AI developers are purchasing millions of physical books, cutting them up, scanning each page into training data, and then discarding the original books.

Confirmed

Global impact / market context

Turning physical books into digital data fuels AI growth but raises copyright concerns and could disrupt traditional publishing, affecting revenue for authors, publishers, and digitization firms.

Analyst inference

The surge in AI model development has increased demand for large text corpora, prompting firms to acquire and digitize physical books at scale, which could affect publishing supply chains and digitization services.

Analyst inference

What to watch

1. Regulators may consider rules on the use of copyrighted physical books for AI training, potentially limiting how firms source training data. Analyst inference
2. Publishers could respond by offering licensed digital versions for AI training, creating new revenue streams or licensing fees. Analyst inference
3. Investors should track companies that provide large‑scale scanning and digitization services, as demand for these capabilities may rise sharply. Analyst inference

Evidence

• decrypt.co