Answer

AI Hallucinations: When Robots Confidently Lie

Sometimes an AI makes up information that sounds totally real but is completely false. It doesn't know it's wrong — it just predicts words that seem like they fit, even when the facts are garbage. Think of it like a student who didn't study but writes a very confident-sounding essay full of nonsense. The AI isn't trying to trick you. It genuinely has no idea it's wrong. That's what makes it dangerous — it sounds so sure of itself.

Example: A lawyer used ChatGPT to find real court cases to support his argument. ChatGPT invented several cases that never existed — with fake names, fake dates, and fake rulings. He submitted them to court and got in serious trouble.