Retrieval-augmented generation, usually shortened to RAG, is a system pattern where an AI model retrieves relevant source material before generating an answer. Instead of relying only on what the model learned during training, the system first looks up documents, then uses those documents as context for the response.
RAG is common in internal knowledge assistants, support tools, and AI search experiences. It is one reason a current, clearly written source page can matter even when a model was trained long before that page existed.
Does RAG guarantee an accurate answer?
No. The retrieved sources can be incomplete, outdated, or misunderstood, and the model can still make a mistake. Good RAG systems show their sources, retrieve relevant material, and give a person a way to inspect the evidence behind an answer.