AI companies accused of hoarding and destroying millions of books

More than a dozen public interest and consumer advocacy groups have called on the Federal Trade Commission to investigate major artificial intelligence developers for alleged practices of purchasing, scanning and destroying books used to train their large language models. The groups describe these actions as “destructive” and “anticompetitive,” arguing that such behavior threatens both public access to knowledge and fair market competition. According to a letter sent to FTC Chairman Andrew Ferguson and Commissioner Mark Meador on Friday, the groups include the Demand Progress Education Fund, Consumer Federation of America and Institute for Local Self-Reliance, among others. They claim that AI companies are buying books in bulk, digitizing the content to train their AI applications and then destroying the physical copies. Some of the books destroyed are reportedly the last surviving copies of original works, raising concerns over the loss of irreplaceable cultural and historical materials. Critics refer to this process as “hoard-and-destroy,” suggesting that it creates barriers for both the general public and smaller competitors seeking to develop their own AI systems. By limiting access to source material, they argue, AI firms may be ensuring that only the wealthiest players can maintain high-quality models. This, they say, results in a monopolization of human knowledge, with the originals being destroyed to facilitate training. The letter also references an August 2024 copyright lawsuit filed by writer Andrea Bartz and others against AI developer Anthropic. The plaintiffs accused Anthropic of acquiring, scanning and discarding millions of print books to train its AI model, Claude. In 2025, a judge ruled that Anthropic’s use of legally purchased books did not violate copyright law. However, the case highlighted ongoing tensions around the ethical and legal boundaries of using copyrighted material for AI training. Separately, a recent investigation by digital publisher 404 Media found that Amazon is engaging in similar activities. According to the report, Amazon is buying books in bulk, scanning them to train its AI tools and then destroying the physical copies. The findings were shared with the public this week, though Amazon has yet to comment on the allegations. These accusations come amid growing scrutiny of the role of AI companies in shaping the future of information access and intellectual property rights. While the FTC has not yet taken action, the pressure from advocacy groups suggests that regulatory attention may soon follow. The groups are requesting that the commission assess the extent of the practice and identify how many of the destroyed books are the final copies of their respective works. The situation raises complex questions about the balance between technological innovation and the preservation of cultural heritage. As AI continues to evolve, so too does the debate over the ethical implications of its development. With more reports emerging, the conversation surrounding the impact of AI on literature and knowledge is likely to intensify in the coming months.