A response generated by a model that sounds coherent and confident but is factually incorrect or made up, resulting from the model predicting plausible text instead of consulting a source of truth.
In GENAI-L (Domains 1 and 3) it is the main risk that justifies the use of grounding and RAG: instead of relying on the model's trained knowledge, the response is anchored to verifiable data to reduce the likelihood of hallucination.
Hallucination (generative AI) is a concept that can show up on the exam. Check the full guide to see it in context.