Several public interest and consumer advocacy organisations have formally requested the U.S. Federal Trade Commission (FTC) to investigate leading artificial intelligence companies for allegedly purchasing, scanning, and destroying millions of books used to train large language models. These practices, described by critics as “hoard-and-destroy,” raise serious concerns over access to cultural materials and potential violations of competition law.
What Happened
On August 21, 2026, a coalition including the Demand Progress Education Fund, Consumer Federation of America, and the Institute for Local Self-Reliance sent a letter to FTC Chairman Andrew Ferguson and Commissioner Mark Meador. The letter accuses AI developers of buying books in bulk, digitizing their contents to train AI applications, and then destroying the physical copies. In some instances, the destroyed books represent the last remaining copies of original works.
The groups warn that this practice restricts public access and limits availability of source materials crucial not only for public knowledge but also for competing AI developers. They assert that such conduct may breach Section 5 of the FTC Act, which prohibits unfair or deceptive acts or practices in commerce.
Key Facts
The letter highlights these key points:
- The alleged practice involves bulk purchasing, scanning, and destruction of millions of print books.
- This “hoard-and-destroy” tactic potentially removes irreplaceable literary works from public access.
- The practice might unfairly block other AI companies from sourcing essential training material.
- The groups have asked the FTC to assess the scale of the destruction and identify how many destroyed books are last copies of the works.
- The letter cites a 2024 copyright lawsuit against AI developer Anthropic for similar behavior; a 2025 ruling found no copyright infringement in Anthropic’s use of legally purchased books to train its AI model.
- A recent investigation by digital publisher 404 Media reported that Amazon is engaged in bulk purchasing and destruction of books for AI training purposes, though Amazon has not commented publicly.
What This Means
If substantiated, these allegations point to a troubling precedent where vital cultural and intellectual property could become increasingly inaccessible to the public and to competing businesses. Book destruction may create artificial scarcity that advantages some AI companies by limiting competition and concentrating control over knowledge resources. This raises broader ethical and legal questions about preserving cultural heritage amid technological advances.
For consumers and researchers, reduced access to diverse original materials risks narrowing the scope and quality of AI training datasets, potentially affecting the accuracy and fairness of AI-generated content. Moreover, this situation challenges the principles of open information and could prompt increased regulatory scrutiny on how data is sourced and used in AI development, affecting the industry’s future direction.
Background
The 2024 lawsuit against AI company Anthropic is a significant antecedent. Plaintiffs alleged that Anthropic’s acquisition, scanning, and destruction of millions of books violated copyright law. However, the court ruled in 2025 that using legally purchased books for AI training did not constitute infringement. This case underscored the complexities at the intersection of copyright, AI, and data usage.
What Remains Unclear
The full scale of the alleged book destruction remains unknown, as do definitive details about the identity and extent of all firms involved. It is also not confirmed how many destroyed books are the sole remaining copies of their works or whether affected authors and rights holders have been notified. The FTC’s response and any ensuing investigation outcomes are pending.
What Comes Next
The FTC has been formally petitioned to open an investigation into these practices. Depending on the findings, the commission could consider enforcement actions relating to unfair competition under the FTC Act. No official announcements regarding the investigation timeline or scope have been made so far.
Sources
This article is based on reporting and publicly available information from the following sources:
Read more AI Regulation stories on Goka World News.