The rapid expansion of artificial intelligence is creating an unexpected dilemma for the literary world. As AI developers search for vast amounts of high-quality training data, millions of physical books are reportedly being purchased, dismantled, scanned, and discarded to build AI datasets.
According to recent reports, specialized suppliers are sourcing books in bulk for AI companies while keeping their clients anonymous. Books published before the rise of generative AI are considered especially valuable because they contain entirely human-written content, making them ideal for training large language models.
The growing demand has also transformed the secondhand book market. Some booksellers have seen sales rise dramatically as AI-related buyers purchase hundreds of titles at a time. While the surge has increased profits, many sellers have expressed concern that rare and out-of-print books are being permanently destroyed during the digitization process.
The issue has sparked debate across the publishing industry. Supporters argue that digitizing books preserves their contents and improves access to knowledge, while critics believe that destroying physical copies erases irreplaceable cultural and historical artifacts that digital files cannot fully replace.
The controversy has also reached the courts. Recent U.S. rulings have found that scanning legally purchased books for AI training can qualify as fair use under certain circumstances. However, AI companies continue to face legal challenges over the use of unauthorized or pirated books, with some firms agreeing to major copyright settlements with authors.
Amid growing criticism, some technology leaders have called for more responsible digitization methods. Elon Musk recently stated that rare books should be preserved and scanned without cutting or destroying their bindings, emphasizing that technological progress should not come at the expense of literary heritage.
As AI continues to evolve, the debate raises a crucial question: How can innovation move forward while ensuring that humanity’s written history is protected for future generations?