AirTag reveals how Amazon destroys rare books for AI training
What happened
Amazon has been purchasing large quantities of printed books, including rare and out-of-print titles, to scan their contents as training data for AI models. The physical books are destroyed after scanning. This practice came to light through information tied to AirTag tracking of shipments and packages.
Why it matters
This exposes how some AI training data comes from the destruction of unique and valuable printed materials. For companies operating in AI, it raises questions about sourcing methods that erode physical cultural assets. For libraries, collectors, and publishers, it signals a pressure point where rare books may be lost not to natural degradation but to extraction for data. Operationally, this reveals a potential supply chain and legal risk tied to how AI training datasets are assembled and the ethics behind destroying physical property for digital gains.
What to watch next
Expect regulatory and public scrutiny targeting data acquisition practices, especially when involving rare or copyrighted books. AI companies may need to justify sourcing methods with more transparency. Watch whether alternatives develop that preserve physical assets or shift to digital-only inputs. For investors and operators, this could impact partnerships, supply chain costs, and the reputational risk profile around AI training data.
AI Quick Briefs Editorial Desk