What is Hallucination?
When an AI model states something false or invented as if it were true.
In plain English
Language models generate likely-sounding text, not checked facts, so they sometimes make up details, sources or numbers. Common defences are grounding answers in trusted documents (RAG), asking the model to say when it does not know, limiting what it may answer, and having a person review important output.
Example
A model cites a court case that does not exist, in a perfectly formal tone.
Easy to confuse with
- Bias in AI: A hallucination is an invented fact. Bias is a systematic unfairness in results.
Where you will meet it
This idea belongs to the subject area of these practice tests. We do not publish or predict real exam questions.
- AWS AIF-C01 practice tests
- Microsoft AI-901 practice tests
- Google Cloud Generative AI Leader practice tests
- IAPP AIGP practice tests
- ISACA AI Fundamentals practice tests
Related terms
- Retrieval-augmented generation (RAG): Letting a model look up relevant documents first and answer using them.
- Responsible AI: Building and using AI in ways that are fair, safe, private, transparent and accountable.
- Large language model (LLM): A very large neural network trained on huge amounts of text to predict and generate language.
Test yourself
Practice tests with an explanation for every answer option, so each mistake teaches you something.
Browse practice testsAll glossary termsWritten and checked by the Be Ready Education team, last reviewed 2 November 2026. General study information, not professional advice. Providers change names and rules, so confirm details on the official documentation. Not affiliated with, endorsed by or sponsored by any certification vendor or exam body. Certification names, exam codes and product names are trademarks of their respective owners and are used only to identify the exam a practice product prepares for. Our questions are original and are not actual exam questions.