Anthropic destroyed millions of print books to build its AI models
1 min read
Summary
AI company Anthropic invested millions of dollars in scanning print books to create an AI system similar to ChatGPT.
This involved cutting and destroying millions of print books in order to train the AI system.
The company hired Tom Turvey, former head of partnerships for Google Books, to lead the scanning operation, which was intended to replicate Google’s successful book-scanning project that had established key fair use precedents.
Judge William Alsup ruled that the scanning operation did constitute fair use of the copyrighted material, due to the company’s purchase of the books, destruction of the print copies post-scanning, and the retention of digital files solely for internal use.
However, Anthropic’s earlier piracy behaviour undermined the company’s position, meaning Anthropic may have achieved the first legally sanctioned case of AI fair use had it adhered to the above strategy from the beginning.
