Microsoft deletes blog telling users to train AI on pirated Harry Potter books

So, here’s a wild one — Microsoft recently pulled a blog post after it sparked a ton of backlash. The post, written by senior product manager Pooja Kamath and shared on her LinkedIn, suggested that developers could train AI models using pirated Harry Potter books. Yeah, you heard that right. The idea was to show how easy it is to add generative AI features using Microsoft tools like Azure SQL DB and LangChain. But here’s where it gets messy — critics called out that promoting piracy isn’t just unethical, it’s a legal minefield. According to Ashley Belanger from Ars Technica, the company quickly deleted the post after the controversy blew up on Hacker News. Now, Microsoft’s trying to clean up the mess, but it’s a reminder that even well-meaning attempts to showcase tech can backfire if they cross the line. And honestly, this raises questions about how big companies handle sensitive topics in their AI development.