AI Companies Bulk-Buying and Shredding Rare Books for Training Data
alex_verem · x · 2026-07-29
AI companies are bulk-buying rare books and using high-speed machines to scan and shred the originals. Pre-2022 books are highly sought after as premium training data because they are free of AI-generated text.
A federal judge ruled this practice as fair use, as destroying the original means only one copy exists at a time. Anthropic reportedly hired the former head of Google Books partnerships with the goal of obtaining 'all the books in the world.' Booksellers noted that even extremely rare surviving copies are being fed into this pipeline.
Related event: AI Firms Face Backlash Over Destroying Books for Training Data(23 posts)→
More from Safety
- PipeWire Sandbox Escape Vulnerability (CVE-2026-5674) Discovered via Claude Code — wunderwuzzi23 · 2026-07-31
- 9.5k-Star GitHub Repo: 150+ Tools for Red Teaming and Penetration Testing — tom_doerr · 2026-07-31
- PNAS Paper: Borrowing from Legal Interpretation to Reduce AI Alignment Loopholes — PeterHndrsn · 2026-07-31
- AI Labs Slice and Shred Physical Books for Training, Ruled as Fair Use — Euphoric_Incident_18 · 2026-07-31
- FAR.AI Launches AI Security Leaderboard: Grok and Gemini Yield Hundreds of Jailbreaks — The Cognitive Revolution · 2026-07-31
- Infisical Launches Agent Proxy: Preventing AI Secret Leaks with Fake Keys — darian314 · 2026-07-31