Tue, 1 Sept 2026
In the News

Secondhand book sales surge as AI training fuels bulk purchases

UnbarNewsUpdated 17 Aug 2026· 2 min read

Used‑book dealers see a sharp rise in large orders, sparking speculation that the volumes are being harvested for artificial‑intelligence datasets and later destroyed.

Secondhand book sales surge as AI training fuels bulk purchases

Secondhand book retailers across the globe are reporting an unexpected spike in demand for bulk lots of used titles. Small independent shops and larger online platforms alike have noticed a surge in orders that often consist of thousands of copies of the same work, far beyond what typical collectors or resellers would need.

Industry observers suggest the pattern points to companies that train large language models. By acquiring physical copies in bulk, these firms can scan and digitise the text, add it to their training corpora, and then discard the originals. Some booksellers have confirmed that after the scanning process the books are sent to recycling facilities, effectively being pulped.

The trend has created both opportunities and challenges for the used‑book market. On the one hand, sellers enjoy higher turnover and better prices for titles that would otherwise sit on shelves. On the other, the sudden influx of bulk orders threatens to distort pricing for rare or out‑of‑print books, as the supply chain becomes saturated with copies destined for data extraction rather than resale.

Environmental groups have raised concerns about the waste generated when scanned books are shredded after use. While the recycling process recovers paper fibers, the loss of physical copies removes cultural artifacts from circulation. Some retailers are now asking buyers to disclose the intended use of large orders, hoping to separate genuine resale activity from data‑harvesting operations.

The phenomenon also highlights the growing appetite for text data in the AI sector. As language models become more sophisticated, the need for diverse, high‑quality training material expands, prompting firms to look beyond publicly available digital sources. The reliance on physical books underscores a gap in open‑access publishing and raises questions about the future of copyright enforcement in the age of machine learning.

While the full scale of AI‑driven book purchases remains unclear, the pattern is reshaping the secondhand market and prompting a debate over the balance between technological progress and cultural preservation.

Original reporting: BBC News.

#books#artificial intelligence#retail#environment#technology