Company that said it could scan and destroy books for AI data-harvesting has deleted that part of its website: ’no such service was ever brought to life'
ISBNdb, a third-party book database company, has deleted its website page advertising services to source printed books for AI language model training. The removal follows reporting by 404 Media about AI companies’ involvement in mass book destruction during scanning for training data—a practice previously obscured through partnerships with third-party rebuyers to avoid negative publicity.
ISBNdb stated that “no such service was ever brought to life,” suggesting the advertised offering never materialized into an actual operational program. The incident highlights growing scrutiny around how AI companies source and utilize copyrighted materials for model training, particularly regarding books and literary works. This practice has drawn criticism from authors, publishers, and copyright advocates who argue that AI developers are benefiting from intellectual property without proper licensing agreements or compensation.
The removal of the service description represents an attempt to distance ISBNdb from the controversial practice, though the company’s advertising had explicitly marketed its ability to supply books “at the scale AI demand” requires. The story underscores the ongoing tension between AI development’s data requirements and copyright protections, as well as the reputational risks companies face when their involvement in ethically questionable practices becomes public.