Literary plunder? The company buying up books and destroying them to train AI

A Canadian company buys old books, scans them and destroys them to train AI. Plunder or recycling? The debate over who controls knowledge in…

English · Original discussion in Spanish · Published

Who is the company buying up old books to feed AI and then destroying them?

A Canadian company has swept into bookshops around the world, buying up thousands of copies with no commercial outlet. According to booksellers, the process is simple: cut off the spine, scan the content and recycle the paper. The result is the disappearance of physical copies that end up feeding an algorithm. The story, published by eldiario.es, has peine the debate over who controls knowledge in the age of artificial intelligence.

The silent loss of cultural heritage

Behind the operation is a company whose identity has not fully come to light. It buys up entire lots of second-hand books languishing in warehouses. The problem is not so much mass-market editions – of which hundreds of copies survive – but local literature, out-of-print texts and one-of-a-kind copies. When the physical medium is destroyed, the information is left solely in the hands of AI. Any future alteration would be impossible to verify. The parallels with “1984” and “Fahrenheit 451” are unavoidable.

Between recycling and digital censorship

Spain publishes more than 90,000 titles a year, and worldwide the figure approaches 2.4 million. Many second-hand books would end up in the bin anyway. The nuance, however, lies in the intent. While conventional recycling poses no threat to memory, digitisation with controlled destruction by a private company opens the door to manipulation. There have already been cases of classics being rewritten, such as Roald Dahl's, or of original titles by Agatha Christie being censored. The difference is that, with paper, an unalterable version existed.

An opaque business in need of regulation

Booksellers are calling for libraries or digital archives such as Archive.org to be consulted before destruction. The technology exists to scan and preserve, but the business of training AI with copyright-protected material is a minefield. The company operates in the shadows, giving no account of which titles it destroys or whether it deposits digital copies. The question that remains is whether this is mere recycling or the beginning of a monopoly on knowledge. With the big tech companies, experience suggests the latter is not so far-fetched.

Also available in: Deutsch

Summary of a discussion on Burbuja.info - Foro de economía, actualidad y política., translated from Spanish and reviewed before publication. Read the full discussion (71 replies).

More summaries

All summaries in English →

-
Back