AI companies are buying up rare books, slicing their spines off, and shredding them to train algorithms, without anyone knowing. Anthropic hired the former head of Google Books partnerships to scan “all the books in the world,” while a service called ISBNdb lets buyers order up to a million books anonymously. A federal judge recently ruled this practice legal under “fair use,” even as rare books, some of the last surviving copies of 18th-century texts, are disappearing into machines that turn them into data.

You’re not just watching this happen. It’s shaping the future of knowledge, ethics, and your business strategy. This article shows you what’s at stake and how to act before the next round of destruction starts.

AI Companies Are Destroying Rare Books, And It’s Happening in Plain Sight

A growing number of AI companies are buying and shredding rare books to train their models. This practice, facilitated by services like ISBNdb, is legal and increasingly common, with serious consequences for cultural heritage. The former head of Google Books partnerships now works for Anthropic, helping them scan “all the books in the world.” A federal judge has ruled the destruction of originals as “fair use,” even as rare books with few surviving copies are being fed into machines that turn them into data. The result is irreversible loss, not just of physical artifacts, but of knowledge that can never be recovered.

How AI Companies Are Destroying Rare Books for Training Data

The Scale of Book Destruction

AI firms are buying and shredding rare books at an unprecedented scale. Services like ISBNdb facilitate bulk orders of up to a million books, with pre-2022 titles being especially valuable because they lack AI-generated text. The process involves high-speed machines that cut spines and shred the originals, reducing centuries-old artifacts to training data. This isn’t a niche practice, it’s growing, with companies like Anthropic hiring former Google Books executives to scan “all the books in the world.”

Rare books that have survived wars, fires, and centuries of handling are now being fed into machines that destroy them for AI training. Some of these books are among the last surviving copies of 18th-century texts, making their destruction irreversible and irreplaceable.

How ISBNdb Enables Anonymity

ISBNdb not only facilitates the purchase of rare books but also ensures anonymity for its clients. The service offers NDAs as a feature, coaching buyers on how to frame the destruction as “digital preservation.” This allows AI companies to operate in secrecy, avoiding public scrutiny or backlash. The company’s website even jokes that “AI company destroys two million books” is not a headline that generates sympathy, highlighting the lack of transparency in the process.

By enabling anonymous transactions, ISBNdb turns the destruction of rare books into a quiet, legal business. This anonymity makes it easier for AI firms to scale their operations without facing immediate consequences or public pressure.

Legal Justifications and Risks

A federal judge recently ruled that shredding rare books for AI training is a form of “fair use,” citing that eliminating the original ensures only one copy exists at a time. This legal framework legitimizes the destruction, allowing AI companies to continue their practices without facing legal barriers. However, this justification raises serious ethical and practical concerns.

The destruction of rare books is not just a legal issue, it’s a risk to cultural heritage. Once a rare book is shredded, it cannot be recovered. This creates long-term risks for knowledge preservation, especially when the books being destroyed are among the last surviving copies of their kind.

Why This Matters to Your Business and Industry

Impact on Cultural Heritage and Data Integrity

The destruction of rare books by AI companies has a direct impact on data integrity and cultural heritage. Once a book is shredded, it’s gone forever. This is not a theoretical concern, rare books with almost no surviving copies are being fed into machines that turn them into training data. The loss of these artifacts affects more than just historians; it impacts the reliability of data sources and the long-term value of knowledge. For industries that rely on accurate, preserved information, this is a growing risk.

Reputational Risks for AI-Dependent Companies

If your business depends on AI, you may be complicit in this destruction without realizing it. Services like ISBNdb facilitate bulk orders of books and offer NDAs to keep the practice quiet. Companies using AI models trained on this data risk being associated with a practice that many view as unethical. As one commenter noted, this is “Burning of the library of Alexandria 2.0.” If your brand is linked to this, it could damage trust and stakeholder relationships.

Opportunities for Ethical AI Leadership

There is an opportunity here for businesses to lead by example. Companies like SpaceXAI are already taking steps to preserve rare books and scan them the “hard way.” Operations leaders and quality managers can push for ethical AI practices that prioritize preservation and transparency. This isn’t just about doing the right thing, it’s about future-proofing your business against the reputational and operational risks of being associated with irreversible data loss.

What You Can Do, Practical Steps to Address the Issue

Advocate for Ethical AI Practices

As an operations leader or quality manager, you have influence over AI adoption in your organization. Push for transparency and ethical sourcing of training data. Demand that AI vendors avoid destructive practices like rare book shredding. This isn’t just about ethics, it’s about long-term risk management. If your AI models are trained on data from illegally destroyed books, it could expose your company to legal and reputational harm.

Support Digital Preservation Initiatives

Encourage your organization to fund or partner with digital preservation projects. These initiatives aim to scan rare books in ways that preserve the original artifacts. Unlike the high-speed shredding machines used by AI companies, these efforts use non-destructive methods. Supporting such work ensures that knowledge is preserved for future generations, and it aligns your company with responsible innovation.

Audit Your AI Vendors for Compliance

Review the practices of any AI vendors you work with. Ask for proof that their training data is ethically sourced. If a vendor is using destructive methods like those described by ISBNdb, it’s a red flag. You can also push for internal audits of AI tools to ensure they don’t rely on data obtained through illegal or unethical means. As one commenter noted, “We are becoming increasingly stupid, careless and thoughtless, everywhere you look.” That mindset has no place in your operations.

The Future of AI and Cultural Preservation

The Acceleration of Book Destruction

With legal loopholes and demand for pre-2022 books still strong, the destruction of rare books is not slowing down. ISBNdb’s business model, built around anonymity and NDAs, ensures the practice continues under the radar. As AI companies scale their training data needs, the pressure on rare book collections will only increase. This is not a hypothetical scenario; it’s a growing industry with real, irreversible consequences.

A Call for Industry-Wide Accountability

Operations leaders and quality managers can’t ignore this. The destruction of rare books is a systemic issue that demands oversight. The former Google Books executive now working for Anthropic is a symbol of where the industry is headed. If no one pushes back, the loss of cultural and historical artifacts will be matched only by the erosion of public trust in AI’s ethical use.

How to Future-Proof Your AI Strategy

Your AI strategy must include ethical sourcing of training data. Avoid vendors that use destructive practices. Demand transparency in how models are trained. The cost of ignoring this is not just reputational, it’s a risk to the integrity of your AI systems. Future-proof your operations by aligning with vendors that prioritize digital preservation and ethical AI development. The alternative is a world where AI gains power at the expense of knowledge itself.

Source: twitter.com