A hallucination is when an AI states something that sounds confident and plausible but is simply wrong. An invented statistic, a feature you don't offer, a product that doesn't exist, or a fabricated source.
For brands this is a real risk, not a curiosity: an engine working from stale or thin information can describe you inaccurately. Wrong pricing, a discontinued product, a capability you've never had, and present it as fact to a buyer. Because the answer reads authoritatively, users rarely question it.
The fixes are the same ones that improve visibility generally: make accurate, current, well-structured information about your brand easy to retrieve, so the engine grounds its answer in your facts instead of guessing from memory. You can't stop hallucination entirely, but you can starve it. The more clearly the truth is reachable, the less room there is for the model to invent.