Hallucination
Definition: A hallucination is an AI answer that is false but stated confidently — the model 'makes up' plausible information.
It's an inherent limit of LLMs: they predict likely text, not verified truth. Hence the importance of citing and checking sources.
Frequently asked questions
Why do AIs hallucinate?
Because an LLM predicts plausible text, not verified truth. Absent reliable information, it "fills in" with what seems likely — and states it confidently.
How do you reduce hallucinations?
Ground answers in sources (RAG), ask for citations, ask precise questions, and verify critical facts. Recent reasoning models also lower the error rate.