Microsoft Removes Blog Suggesting AI Training on Pirated Harry Potter Books

Microsoft Removes Blog Suggesting AI Training on Pirated Harry Potter Books

Feb 20, 2026, 12:11 PM en 11 views

TL;DR

Microsoft removed a blog post that recommended training AI models using pirated Harry Potter books marked as public domain. The blog faced backlash for potential copyright infringement.

Following criticism on Hacker News, Microsoft deleted a blog that advised using a dataset of Harry Potter books for AI training, incorrectly marked as public domain. Written by Pooja Kamath, the post was intended to showcase a new Azure feature by using popular literary works. This raised concerns about copyright infringement. Microsoft did not comment, and the dataset has since been removed from Kaggle.

Sources & References

More in Technology