Amazon's North Las Vegas Facility Destroys Books for AI Training

https://www.youtube.com/watch?v=PTMxpzHtisI

Amazon is facing criticism after an investigation found that books are reportedly being cut apart, digitally scanned and discarded at a North Las Vegas facility as part of efforts to obtain data for artificial intelligence training.

The revelation has sparked questions about copyright, ethical AI development and the preservation of rare books, particularly because some of the books involved are described as difficult to replace.

Rare Books Tracked to Amazon Facility

The investigation by 404 Media began after a rare-book seller agreed to place an Apple AirTag inside one book from a shipment of roughly 1,000 books suspected of being purchased for AI training.

The tracking device eventually led investigators to an Amazon facility in the Las Vegas area, identified as VGT3 within the larger LAS8 complex. Employees told 404 Media that their work involved cutting the bindings off books and running the individual pages through scanners.

Once the books are cut apart and scanned, the physical copies are reportedly discarded.

Why Are the Books Being Destroyed?

AI companies need enormous quantities of high-quality text to train large language models. Rare, older and out-of-print books can provide material that may not be readily available online.

The destructive scanning process allows pages to be separated and fed through industrial scanners more quickly, but it permanently destroys the physical book.

Amazon told 404 Media that it purchases books through commercial channels to help develop and improve its products and services.

Copyright Questions Remain

The practice has also raised questions about whether companies can legally acquire copyrighted books, digitize their contents and use that material to train AI systems.

The legal landscape surrounding AI training and copyrighted works remains complicated. Recent court proceedings involving AI companies have intensified the debate over fair use, licensing and the use of copyrighted material to develop AI models.

More than a dozen civil society organizations have also urged the Federal Trade Commission to investigate the practice, arguing that the acquisition and destruction of books could raise broader competition concerns in addition to copyright issues.

Concerns Over Losing Rare Books

For booksellers and collectors, the physical destruction is particularly concerning when the books are scarce or difficult to replace.

Some sellers have reportedly become reluctant to fulfill large orders when they suspect the books are being purchased for AI training. While a book may have limited commercial value, destroying one of only a small number of surviving copies can permanently remove an important piece of literary or historical material.

The controversy highlights a growing tension between the demand for AI training data and the preservation of human-created works.

As AI companies continue searching for new sources of high-quality information, the debate over how that material should be obtained—and what happens to the original works afterward—is likely to continue.