AI companies buy millions of books, scan them for training data, then discard originals

AI developers are quietly purchasing physical books by the millions, slicing them apart, scanning every page into training datasets, and then discarding the originals. The practice is carried out without the knowledge or consent of authors or publishers. Critics have compared the phenomenon to book burning, pointing to the mass-scale destruction of cultural heritage.

More on this topic

💻 Technology AI firms secretly buy and destroy millions of books via middlemen to avoid scrutiny PC Gamer · 1d ago 💻 Technology Authors sue AI companies over training data — and some are winning The Verge · 1d ago 💻 Technology Anthropic's $1.5bn Book Piracy AI Settlement Approved by Court The Verge · 8d ago

More on this topic

💻 Technology AI firms secretly buy and destroy millions of books via middlemen to avoid scrutiny PC Gamer · 1d ago 💻 Technology Authors sue AI companies over training data — and some are winning The Verge · 1d ago 💻 Technology Anthropic's $1.5bn Book Piracy AI Settlement Approved by Court The Verge · 8d ago