Hallucination (AI)
When an AI confidently produces false or fabricated information.
A hallucination is when an AI model generates information that sounds plausible but is incorrect or invented. Because the output is fluent and confident, hallucinations can be hard for users to spot.
Good AI product design mitigates this with grounding (RAG), visible sources, confidence cues, and editable outputs — designing for the reality that the model is sometimes wrong.
Key characteristics
- When an AI model produces confident but false or fabricated information.
- Stems from generating plausible text from patterns rather than verified facts.
- Especially risky for citations, statistics, quotes and recent events.
- Mitigated with grounding (RAG), source citation and human review.
Example
An LLM invents a realistic-looking but non-existent research paper and citation when asked for a source — a classic hallucination that must be fact-checked.
Frequently asked questions
Why do AI models hallucinate?
They predict likely text rather than retrieving verified facts, so when they lack grounding they fill gaps with plausible-sounding inventions, stated just as confidently as true information.
How can you reduce AI hallucinations?
Ground responses in trusted data with RAG, ask for sources and verify them, lower randomness, constrain the task, and keep a human in the loop for anything factual or high-stakes.
Need this done, not just defined?
I design and build it — Framer sites, product UI/UX, and more.