AI firms destroying millions of rare books for training data, raising alarms over cultural heritage

Major AI developers are reportedly acquiring and destroying vast quantities of physical books to create training datasets. A recent court ruling deemed this practice "fair use," sparking concerns about the preservation of cultural artifacts.
Key takeaways
- AI companies are destroying books for training data.
- Court ruling supports destructive scanning as fair use.
- Concerns grow over cultural heritage preservation.
- Data sourcing impacts future access to physical texts.
Why it matters
This practice raises questions about the long-term availability of physical texts for research and historical study. Users of AI tools should be aware that the data underpinning these systems may come at the cost of cultural heritage.
Try this on SynaBot
Related AI assistants, prompts, and tools from the SynaBot catalog.


