Amazon: Scandal over the destruction of rare books for artificial intelligence training

Recent investigative reporting has uncovered disturbing facts about the activities of e-commerce giant Amazon. According to 404 Media, the company systematically buys rare books that are later destroyed in the process of training artificial intelligence systems. This shocking case raises questions about the ethical standards of doing business in the digital age and about the value of cultural heritage.

AI’s digital “appetite”: how rare editions become victims of progress

The investigation gained momentum after the 404 Media team used a clever tracking method. They embedded an Apple AirTag in one of the rare books, which was part of a large order consisting of a thousand similar copies. The book, it turned out, ended up at an Amazon warehouse in Las Vegas, located in a facility with the code VGT3.

Journalists’ sources working at this warehouse confirmed the disturbing procedure taking place inside. According to them, the first step is to cut the spines off the books, which effectively destroys their integrity. After that, the books are sent through a scanner. It later emerged that the VGT3 warehouse is only part of a larger logistics center in Las Vegas known by the code LAS8. The main duty of workers at this site is precisely the process of “demolishing” books for further scanning.

Even the logo associated with VGT3 features an image of a dinosaur allegedly ready to tear a book to pieces, symbolically reflecting the cruelty of the process. This investigation began in July, when 404 Media journalists, working with a seller, arranged to place a tracker in one of the books included in the large batch.

An unexpected surge in demand and the role of online marketplaces

The very fact of such a large order for rare books already raises suspicion. However, in the context of the rapid development of artificial intelligence technologies, a paradoxical trend is emerging: demand for books is rising. The reason is the active buying up of books by major tech companies that use them to train their AI models.

The order that became the focus of the investigation was placed through Biblio, an online platform that presents itself as the world’s largest independent book marketplace. It is important to note that such bulk orders often include a wide variety of books. For example, one bookstore in Ireland reported receiving an order for 5,000 books, which management suspects are intended for the same purpose — AI training. Among them were both high-quality nonfiction titles, such as The History of Connemara, and, presumably, less valuable but still books such as Yeadi Hobbs’s Guide to SSIA.

ISBN as the key to AI’s digital library

404 Media journalists claim that Amazon warehouse employees, such as those at VGT3, are required to scan the unique ISBN code of each book. This supports the theory that companies developing artificial intelligence are creating complete digital archives of every book ever published, or at least of every book with an ISBN. One bookseller noted that books without ISBNs never appear in such large orders, indicating a systematic approach to selecting materials.

Previous cases and the broader context of the problem

This Amazon case is not the first, and likely not the last of its kind. Earlier this year, court documents revealed information about Anthropic’s project called “Panama.” This project, planned for 2024, aimed to fully scan all books in the world while destroying them. Although the court allowed Anthropic to use books to train its own AI models, the company was fined $1.5 billion for storing 7 million pirated copies of books.

Another high-profile court case last year found that Meta had illegally scanned 82 terabytes of book content for its own needs. These cases highlight the scale of the problem and suggest that using copyrighted materials to train AI without proper permission and compensation is becoming a widespread practice.

Cultural value versus technological progress

The situation with Amazon raises important questions about the balance between technological progress and the preservation of cultural heritage. On one hand, the development of artificial intelligence promises major benefits for humanity, from scientific breakthroughs to improvements in everyday life. However, the methods used to achieve these goals can have destructive consequences.

Rare books are not just paper pages, but carriers of history, knowledge, ideas, and human experience accumulated over centuries. Their destruction, even in the name of progress, is an irreparable loss for future generations. It calls into question the very notion that progress justifies any means.

Next steps and the role of society

The 404 Media investigation seriously calls into question the reputation of Amazon and other tech giants. This news has sparked outrage among authors, publishers, booksellers, and anyone who values books as cultural artifacts.

Companies developing AI must act transparently and ethically. That means not only obtaining permission to use materials, but also seeking alternative, less destructive training methods. It may be worth considering the creation of specialized digital archives that preserve books in their original form while still allowing them to be used for AI training.

Society must also play an active role in this debate. Public pressure, awareness, and support for initiatives aimed at protecting copyright and cultural heritage can help shape a more responsible approach to technological development. This Amazon scandal is a warning sign that demands immediate attention and action to ensure that technological progress does not come at the expense of destroying our cultural memory.

Economic aspects and market conditions

When analyzing the situation, the economic dimension cannot be ignored. The massive volumes of book buying point to enormous demand from those building the future on artificial intelligence. This demand creates new markets and opportunities for booksellers, especially those specializing in rare and antique editions. However, it also creates risks for the preservation of unique editions, which may be swallowed up by this “digital appetite.”

The existence of platforms like Biblio, which act as intermediaries in such large deals, underscores how the book market is being transformed. Previously, such orders would have been almost impossible, but now, thanks to online technologies, they are becoming a reality. This also requires such platforms to take greater responsibility and pay closer attention to the source of books and their ultimate fate.

Intellectual property in the digital age: challenges for legislation

This Amazon incident once again brings the long-standing issue of intellectual property in the digital age to the forefront. Laws governing copyright often fail to keep pace with rapid technological development. Companies developing AI may find loopholes in existing laws to use protected materials.

The lawsuits against Meta and Anthropic show that the legal system is trying to adapt, but the process is slow and difficult. Legislators at the international level must develop new, more effective mechanisms for protecting copyright that take into account the realities of AI training. This could include new licensing models or stricter penalties for unauthorized use of materials.

Conclusion: an ethical choice at the crossroads of progress and heritage

The scandal involving Amazon and the destruction of rare books is only the tip of the iceberg. It exposes deeper problems related to the ethics of data use, the preservation of cultural heritage, and legal regulation in the digital era. Every company striving for innovation must make a conscious choice: are they willing to put our history and cultural heritage at risk in the name of technological progress? A future built on knowledge should not be built on the destruction of its original sources.