Hallucination

What it means. A hallucination is when an AI model states something false as if it were true. The model is not lying. It is predicting plausible text, and sometimes the most plausible-sounding answer is simply wrong.

Why it matters. Hallucinations look exactly like correct answers. They come with the same confident tone, so a user cannot tell the difference without checking. In business use, that is a real risk.

What beginners often misunderstand. Hallucination is not a bug that gets fully patched out. It is a property of how these models work. You reduce it with grounding, sources, and verification, but you do not assume it is gone.

How it shows up in real projects. Made-up citations, invented product features, wrong prices, fake policies, and confident answers about data the model was never given. The fix is design: give the model real sources and check important outputs.