AI Glossary
What is an AI hallucination?
Hallucination: A hallucination is when an AI states something false with full confidence — invented facts, fake citations, wrong numbers, non-existent court cases.
In plain English
An LLM's job is to produce plausible text, not verified truth. It is like a brilliant student who would rather guess fluently than say "I don't know" — the guess sounds exactly like a correct answer, which is what makes it dangerous.
Why it matters
- Always verify names, numbers, dates, quotes and links before using them
- Hallucinations get worse on niche topics where training data was thin
- Asking the model to cite sources, or grounding it with RAG, reduces (not eliminates) the problem
A concrete example
Lawyers have been sanctioned for filing briefs citing court cases an AI invented — complete with realistic case numbers. The citations looked perfect; none existed.
Related terms
Learning AI from scratch? Start with the free Learn AI course picks, then put it to work with 50 tested prompts — or get the daily 2-minute briefing, free →