Google documents Google-Extended token to opt out of Gemini training

What happened

Google's crawler documentation defines Google-Extended as a standalone robots.txt product token that web publishers can use to control whether content Google crawls may be used for training future generations of Gemini models (powering Gemini Apps and the Vertex AI API for Gemini) and for grounding in those products. Google states that Google-Extended "does not impact a site's inclusion in Google Search nor is it used as a ranking signal" — so a site owner can opt out of Gemini training without affecting Google Search visibility. Unlike a crawler, Google-Extended has no separate HTTP user-agent string; it works as a control token in robots.txt.

Why it may matter to you

For a creator with a website, this is the followable opt-out Google offers for Gemini training: add a Google-Extended disallow rule to robots.txt, e.g. User-agent: Google-Extended followed by Disallow: /. It preserves Google Search inclusion and ranking, which is the trade-off site owners usually want. It does not cover content already crawled and used. The documentation does not address whether it affects AI Overviews or AI Mode.

What is not yet known

The documentation does not say how quickly the token takes effect in training pipelines or offer a verification mechanism. Because Google-Extended is a permission token rather than a crawler, compliance is a matter of Google's stated policy rather than something observable in server logs.