Concept

Hallucination (AI)

When an AI system states something false or made up as if it were true, such as fake quotes, sources or numbers.

Last reviewed · 2 sources · Suggest a correction

Contents
  1. In one line
  2. Why it happens
  3. How to protect yourself
  4. Sources

In one line

A hallucination is a confident, fluent answer that is wrong. The US National Institute of Standards and Technology (NIST) uses the term confabulation for the same risk.

Why it happens

A large language model generates the most plausible-sounding text, not verified facts. When it lacks information, it can fill the gap with something that sounds right: an invented court case, a book that doesn’t exist, or a wrong statistic.

How to protect yourself

  • Ask for sources, then open them. A real source should exist and say what the AI claims.
  • Be most careful with legal, medical, financial and news questions.
  • Treat AI answers as a first draft, not a final answer.

This is also why AI detectors and summaries can be wrong about you. See what to do if you are falsely accused of using AI.

Sources

  1. Artificial Intelligence Risk Management Framework: Generative AI Profile (NIST AI 600-1), NIST
  2. Introduction to large language models, Google for Developers

Every fact on this page comes from the sources above. Entries are re-checked when things change. Spotted something out of date? Tell us. Read our editorial policy.