Hallucination

A phenomenon where a Generative AI model or Large Language Model generates output that is factually incorrect, nonsensical, or unaligned with its training data.

Part of the Data & AI Terms glossary.