What is AI hallucination?

An AI hallucination is when a model states something false or made-up as if it were fact — fluent, confident and wrong.

Language models generate plausible-sounding text rather than looking up verified truth, so when they don't know something they can fill the gap with an invention — a fake citation, a wrong figure, a quote nobody said — delivered in the same assured tone as a correct answer. That confidence is what makes hallucinations easy to miss.

You can't eliminate them, but you can cut them down: give the model the source material to work from instead of relying on its memory, ask it to say when it's unsure, request citations you can check, and keep temperature low for factual work. The reliable habit is to treat anything checkable — names, numbers, quotes, dates — as a draft to verify, not a fact to trust.