Back to the AI glossary Basics
Hallucination
When AI generates convincing-sounding but false information.
Explanation
A hallucination occurs when an AI model invents facts that sound plausible but are not true. This happens because the model is based on probabilities - not truth.
How it works
LLMs always generate the "most likely continuation". If the model doesn't know a real answer, it fills the gap with plausible-sounding text - without knowing that it's wrong.
Example
A chatbot invents a paragraph number in a law that doesn't exist or cites a non-existent research paper.
Why it matters
Hallucinations are the biggest trust risk when using AI. Countermeasures such as grounding and RAG are therefore essential.