AI and machine learning
RAG
Short for retrieval-augmented generation.
- AI and machine learning
Definition
Short for retrieval-augmented generation. A technique that gives an AI model relevant documents or trusted sources to look up before it answers, so its response is grounded in real content rather than just its training.
Example
A help assistant retrieves the latest approved policy document before drafting an answer and citing the relevant passage.
See how it works
- Person asks a question
- System retrieves relevant documents
- Relevant passages are added as context
- Model drafts an answer for verification
Check your understanding
Common mistake: Retrieval does not automatically make an AI answer correct. Retrieved documents may be outdated, irrelevant or malicious.
Does RAG mean the model has permanently learned the retrieved document?
Show explanation
No. Retrieved context is typically supplied for that request rather than changing the model's trained weights.
For your role
Choose your role to see what this term means in your work.
Why it matters
It helps AI give answers based on your own trusted information.
Where it is used
AI assistants built on company knowledge.