RAG (Retrieval-Augmented Generation)
RAG, short for retrieval-augmented generation, is a way of building AI tools so they look things up in an outside collection of text before they answer. The tool pulls in passages, which came from Wikipedia in the original 2020 paper by Patrick Lewis and colleagues, then writes its answer from those passages plus what it learned in training.
Source: Lewis and others, Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks (2020), checked October 29, 2026
What RAG (Retrieval-Augmented Generation) means for your business
OpenAI says ChatGPT may search the web on its own when current information would help, and you can also turn search on from the tools menu. When it uses outside search providers, OpenAI says it typically rewrites your question into one or more searches. Google says AI Overviews and AI Mode may run several related searches to build an answer. To us, that's RAG with the open web as the collection.
For a business, a lookup is the moment your pages can be found and quoted. We'd make sure AI crawlers can reach your pages, and we write each section so it still makes sense if it's the only part an AI tool reads.
If ChatGPT searches your category and quotes a competitor's page instead of yours, building pages that get quoted is the work we'd do first.
Last reviewed October 29, 2026. All terms
How AI answers it today
On October 29, 2026, Google read the bare acronym as the AI term only, and its AI Overview cited cloud companies' explainers, Wikipedia, a university page, Reddit and YouTube. ChatGPT explained it with a company-documents example and cited no sources.
Our run through ChatGPT and Google's AI Overview, October 29, 2026. A sample, not what every buyer sees.
Related terms in answers and how content gets used
Find out where your brand stands.
Tell us what you're trying to figure out. We'll get back to you within one business day with a time to talk.