In more detail
Google-Extended isn't a separate crawler. Google's documentation describes it as a standalone product token: crawling is still done by Google's normal user agents, and the token is only used to read your preference. That's why you won't see "Google-Extended" in your server logs.
According to Google, the token controls whether content it crawls from your site may be used:
- for training future generations of Gemini models that power the Gemini apps and the Vertex AI API for Gemini, and
- for grounding, which means supplying content from the Google Search index to the model when it answers, in the Gemini apps and Grounding with Google Search on Vertex AI.
Google is explicit that Google-Extended "does not impact a site's inclusion in Google Search nor is it used as a ranking signal". AI Overviews and AI Mode are part of Google Search, so they're governed by Googlebot and the normal snippet controls (nosnippet, max-snippet, noindex), not by this token.
To opt out, add a group for it to robots.txt:
User-agent: Google-Extended
Disallow: /
Apple has a similar control token, Applebot-Extended, for its AI models.
Why it matters when hiring an agency
There's a common misunderstanding that blocking Google-Extended keeps a site out of AI Overviews. It doesn't, and it won't hurt your rankings either. If an agency recommends blocking or allowing it, they should be able to explain exactly what changes, and what doesn't. Our AI crawler access checker shows how your robots.txt treats Google-Extended alongside other AI bots.