Amazon is buying and destroying rare books, cutting off their spines, and scanning them for AI training, according to a report that placed a tracking device in a rare book that ended up at an Amazon facility.
Amazon’s transformation from a bookseller to an AI data collector has sparked controversy after a report revealed the company is dismantling rare volumes to harvest their contents for machine‑learning models.
How the practice was uncovered
TechCrunch reported that a tracking device was covertly placed in a centuries‑old book, which later surfaced at an Amazon fulfillment center. The device’s movement confirmed the book had been taken apart, its spine cut, and its pages digitized.
The investigation noted that the book, described as a "rare" text, was handled without any preservation measures, raising concerns among librarians and historians about the loss of cultural heritage.
Amazon’s stated rationale
Amazon claims the scanning of such books is part of its effort to build more accurate large‑language models, arguing that a broader corpus improves AI performance and benefits users across its ecosystem.
The company’s spokesperson emphasized that the process complies with copyright law and that the digitized content is used solely for internal research.
Reactions from the preservation community
Experts warn that destroying rare books for AI training is a form of cultural vandalism. "When a unique artifact is dismantled, the knowledge it holds is lost forever," said a representative of the Rare Book Preservation Society.
- Loss of physical artifacts that cannot be replaced
- Potential violation of donor or acquisition agreements
- Erosion of trust between libraries and corporate entities
We are witnessing a clash between commercial AI ambitions and the stewardship of humanity’s written heritage.
The incident has prompted calls for clearer regulations on how private companies may access and process rare and fragile cultural materials.