Governance
What does Hallucination mean?
A hallucination occurs when a language model produces information that sounds convincing but is factually wrong or entirely made up. The cause is that models calculate probabilities for the next piece of text – they do not verify truth.
In a business setting this is a key risk. Countermeasures include RAG (answers grounded in verifiable sources), source citations and a human in the loop who signs off on critical results.
Related terms
Retrieval-Augmented Generation (RAG)
A method where a model retrieves relevant documents before answering and grounds its response in them.
Human in the Loop
A principle where people review and approve a critical decision made by an AI system.
Large Language Model (LLM)
An AI model trained on large amounts of text that understands and generates language.
Ready to hand off the routine work?
In a short conversation we find the use case with the biggest leverage in your business – with a clear goal and a measurable result.