AI companies are acquiring and destroying physical books after digitizing them, raising concerns over the loss of rare titles.
AI Companies Engage in Destructive Book Practices
In a concerning trend, AI firms are reportedly acquiring physical books, scanning them for training purposes, and then destroying the originals. This practice raises alarms among booksellers and preservationists, especially as demand for rare and out-of-print titles surges.
According to a report by 404 Media, these companies are utilizing third-party intermediaries to discreetly procure books in bulk. The intermediaries claim to have access to hundreds of thousands of titles, emphasizing the secretive and controversial nature of their operations.
Critics of this trend argue that AI companies are driven by a desire to secure high-quality, human-generated knowledge before it is overshadowed by AI-generated content, often labeled "AI slop." Books published before the rise of generative AI in 2023 are particularly sought after for their superior quality.
The impact on the second-hand book market has been notable. One anonymous bookseller reported an increase in weekly sales from about 20 to several hundred books after AI buyers began purchasing titles. While this has proven financially beneficial, the seller expressed concern about the irreversible loss of unique and rare works once they are digitized and destroyed. "I dislike the end result, and I am concerned about the fate of rare books being pulped," he shared with 404 Media.
This destructive scanning aligns with practices seen in Anthropic's "Project Panama," which similarly involved digitizing millions of books. In a significant legal ruling, a federal judge determined that the digitization of legally owned physical books, even with their destruction, could be considered transformative fair use. This decision has been supported by subsequent rulings in copyright cases involving other AI companies like OpenAI and Meta.
However, not all developments in this arena are positive for AI firms. Anthropic recently faced a $1.5 billion settlement due to the unauthorized use of pirated books for training its AI model, Claude, requiring compensation to affected authors. In response to the growing backlash, some AI developers, including Elon Musk, have publicly opposed the destruction of books, advocating for traditional scanning methods that preserve the originals. Musk stated, "I’ve instructed the SpaceXAI team to safeguard any rare books in a library and to scan them using traditional methods instead of simply cutting off the spine."
