Grounding: definition

Grounding is tying a model’s answer to a verifiable source, such as your documents or live data, so it reports facts instead of inventing them.

An ungrounded model answers from its training and its guesswork. A grounded one answers from specific, citable sources you provide, which is the single most effective defence against hallucination.

RAG is the common way to ground a model, but grounding is the broader principle: never let the system state a fact it cannot point to. We treat it as a requirement for anything client-facing.

Related terms

Retrieval-augmented generation (RAG): Retrieval-augmented generation is a technique that fetches relevant passages from your own documents at question time and hands them to the model, so its answer is grounded in your data instead of its training.
Hallucination: A hallucination is when a model states something false or invented as if it were fact, because it predicts plausible-sounding text rather than looking anything up.
Knowledge base: A knowledge base is the curated collection of documents an AI system draws on to answer questions, and keeping it accurate and current is what keeps the answers accurate and current.