What is retrieval-augmented generation (RAG)?
Also known as: RAG
Retrieval-augmented generation (RAG) is a technique where an AI system retrieves relevant documents from a knowledge source at query time and feeds them to a language model, so answers are grounded in specific, current content instead of the model's training alone. It is how most enterprise assistants answer from company data. From a governance view, RAG means that whatever the retrieval step can reach is what the AI can expose.
RAG is the pattern behind most assistants that answer from your own content: retrieve the relevant material, then let the model compose an answer from it. It is why an assistant can cite a specific internal document rather than hallucinate.
For governance, the important part is the retrieval step. It runs with some identity and scope, and it can only pull what that scope allows. So the controls that matter for a RAG system are the same access controls that govern the underlying content: who and what the retrieval can reach is exactly what the AI can surface.
