What is AI hallucination?
When an AI confidently states something false — a fake fact, quote, source or statistic.
Language models generate the most plausible-sounding text, not verified truth. When they lack information they can fill the gap with invented details that sound right: a book that doesn't exist, a wrong date, a made-up court case.
You can reduce hallucinations by asking the AI to cite sources, turning on web search, giving it the document to work from, and telling it to say “I don't know” when unsure. For anything that matters — health, money, law, exams — verify it yourself.
Spot the warning signs
Very specific numbers or quotes with no source, links that don't open, and names you can't find in a search are classic hallucinations.
Related terms
The type of AI behind chatbots like ChatGPT, Gemini and Claude, trained on huge amounts of text to predict and generate language.
A technique where the AI first looks up relevant documents and then answers using them, so replies are grounded in real sources.
An AI model that works through a problem step by step before answering, trading speed for accuracy on hard questions.
