I work in archive management at my uni, this is standard practice and can easily be undone by rebinding the book. It’s been used for years to digitize books.

12

I did some reading on this and here’s what I found out: The “destroying” part is both slicing the binding to scan pages quicker, as well as shredding the books after they’ve been scanned. The purpose is get clean data from professionally edited books, that have zero AI generated content. The books in question are primarily second hand, as the First-Sale Doctrine allows companies to legally “pirate” the book.

4

exactly

1

The shredding is also a legal loophole, where AI companies can argue that they are using the books for internal model training, or a “Transformative use” under fair use laws.

1