AI firms are buying up old books because they are the last slop-free data left

The Next WebPublished Jul 22, 2026

ISBNdb supplies pre-2022 printed books to AI labs for training data to avoid AI-generated text contamination and data poisoning. The process involves destroying books by removing spines for scanning, with confidentiality agreements to keep buyers anonymous.

#ai-training-data #data-poisoning #isbndb #model-collapse #pre-2022-books

AI firms are buying up old books because they are the last slop-free data left

More in Machine Learning

PyTorch Conference 2026: Google and Meta Open-Source TorchTPU for Native PyTorch on TPUs

Development of an automated method for audiometric data extraction and analysis using voting optical character recognition

Hugging Face Releases Carbon-X, a 14B Open DNA Language Model that Brings DNA and Natural Language Together for the First Time