AI Firms Are Still Buying Out-of-Print Books by the Pallet and Destroying Them After Scanning
Reporting resurfacing across Hacker News (739 points, 468 comments) and mainstream outlets on July 27 documents an ongoing practice: bulk purchases of rare and out-of-print books, destructive scanning, then pulping. ISBNdb reportedly brokers anonymous orders from 1,000 up to a million titles, with buyers prioritizing pre-2022 publication dates specifically to avoid AI-generated text contaminating training corpora. The legal cover is Judge William Alsup's June 2025 ruling in Bartz v. Anthropic, which held that digitizing legally purchased print books for LLM training is fair use; Anthropic's internal 'Project Panama' planning document described the goal as 'destructively scan all the books in the world.' Booksellers quoted in the coverage note the obvious asymmetry — the last surviving copies of an 18th-century text are not replaceable once shredded.
Source
↳ Follow the thread