What is an AI hallucination?

An AI hallucination is a statement a language model produces confidently that isn’t supported by its sources or by reality, such as a claim a paper never makes or a citation to a paper that doesn’t exist. It happens because models generate likely-sounding text; grounding answers in retrieved passages and citing them reduces it and makes it easy to catch.

Invented references are the classic case in research: plausible authors, title and journal, no such paper.

The fix is verification: open the cited passage, and resolve every DOI.