How AI engines work

How do Google AI Overviews pick their sources?

Short answer
AI Overviews run on Google's core ranking and quality systems, then select passages through query fan-out — breaking one question into several sub-queries and retrieving a passage for each. Because selection happens at passage level, most cited pages do not rank in the organic top ten for the original query.

Passage selection, not page ranking

This is the single most important thing to understand about AI Overviews. Google is not summarising the top ten results; it is assembling an answer from passages retrieved across a fan of related queries. A page can be cited without ranking well for anything you were targeting.

The practical consequence is that narrow, specific pages punch far above their weight. A page answering one sub-question thoroughly can be cited in an Overview for a much broader query it could never rank for.

Google's own position

Google's guidance is that optimising for AI Overviews is still SEO — the same crawlability, quality, and helpfulness systems apply. There is no separate ranking pipeline to game.

What changes is the unit of competition. Instead of one page per keyword, you want coverage across the cluster of sub-questions a topic fans out into, each answered cleanly enough to be extracted on its own.

See where you actually stand

We ask AI engines the questions your buyers ask and show you whether you were named, who was named instead, and which sources they cited. Free.

Related questions

More on how ai engines work

Last reviewed 2026-07-30.