Answers ยท Fundamentals

How does Gemini choose sources for legal questions?

Google has not published a formula for which sources Gemini cites. Its own overview says Gemini relies on external sources such as Google Search, is grounded in Google's understanding of authoritative information, and ranks candidate responses by quality. For a law firm, that points to two things you control: whether Google may use your pages for Gemini at all, and whether a page answers the exact question a client asked.

What Google has said

Google's overview of how Gemini works, last updated July 25, 2024, says Gemini relies on external sources such as Google Search, a process it calls retrieval augmentation. It describes a check that filters out harmful or offensive responses, after which the remaining responses are ranked by quality. The same page warns that Gemini can confidently generate responses that contain inaccurate or misleading information.

Google's developer documentation for the Gemini API describes a similar pattern for grounding with Google Search: the model analyzes the prompt, decides whether a search would improve the answer, and if so runs one or more queries before writing a response with citations. That page describes the API, not the consumer app, so treat it as a guide to the general approach rather than a promise about any one answer.

On the reader's side, the Gemini Apps help page on sources says Gemini sometimes shows sources and related content, that not all responses include them, and that when Gemini directly quotes a large amount of text from a webpage, that page appears in the sources list.

The setting that decides eligibility

Google's crawler documentation describes Google-Extended as a product token publishers can use to manage whether content Google crawls may be used to train future Gemini models and for grounding in Gemini Apps. Google says it does not affect a site's inclusion in Google Search and is not a ranking signal there. A firm that blocks Google-Extended in robots.txt stays in Search but tells Google not to use its pages for those Gemini uses. Our answer on which AI crawlers to allow covers the other bots.

Google does not say how Gemini weighs one legal page against another, so this is a working hypothesis from testing: when Gemini searches, a page that answers the question directly, names the state and court, and cites the statute is easier to ground an answer in than a general overview. The family law guide shows that kind of page. Whether Gemini then links you or only names you is a separate measure, covered in citations versus mentions.

Go deeper

The GEO guide for law firms covers how each assistant finds sources and what makes a legal page worth citing.

Common questions

Does blocking Google-Extended remove my firm from Google Search?

No. Google says Google-Extended does not affect a site's inclusion in Google Search and is not used as a ranking signal there. It governs whether Google may use your content for Gemini model training and for grounding in Gemini Apps.

Keep reading

Sources

  1. Google, "What is Gemini and how it works" (last updated July 25, 2024) (linked above)
  2. Gemini Apps Help, "View related sources from Gemini Apps" (linked above)
  3. Gemini Apps Help, "Learn about responses from Gemini Apps"
  4. Google for Developers, "Grounding with Google Search," Gemini API (last updated October 13, 2026)
  5. Google Crawling Infrastructure, "Google's common crawlers" (last updated October 15, 2026) (linked above)

See one of your pages rewritten for AI search.

Send us a page from your site. We email back a rewritten version you can keep. A call is optional.