An investigation by 404 Media revealed that Amazon purchases printed books in bulk, scans them for AI training data, and then destroys the original copies. The process was tracked using an AirTag placed inside a shipment of rare books, which ultimately arrived at an Amazon warehouse in Las Vegas. Workers at the facility confirmed that they cut off book spines to expedite scanning, effectively destroying the books in the process. Amazon uses the scanned data to train its Nova models, according to a spokesperson, who stated the company buys books through commercial channels to improve its products. Booksellers suspect that AI companies are systematically scanning every book by ISBN number, as printed texts often lack online versions and predate 2022, making them free of AI-generated content.
Both Amazon and Anthropic have been involved in similar practices, with Anthropic facing a lawsuit over its 'Project Panama.' The lawsuit revealed that Anthropic purchased books from marketplaces, cut off their spines, and digitized them. A judge ruled that the scanning qualified as fair use and did not violate copyright, partly because the printed originals were destroyed and not resold. This practice has sparked controversy, as it turns sometimes rare books into private training material, removing knowledge from public access and locking it inside corporate AI models.
Source: thedecoder
