A robots.txt control that lets sites opt out of having their content used to train and ground Google's Gemini models.
ID
SS-DF-109
Confidence
Established · 76
Evidence
Established
Updated
2026-07-08
Definition
A robots.txt control that lets sites opt out of having their content used to train and ground Google's Gemini models.
Google-Extended is a separate user-agent token that governs whether your content feeds Google's generative AI, independently of normal Googlebot crawling for Search. Disallowing it can withhold your pages from Gemini training and grounding while leaving classic search indexing intact. It gives publishers a lever over AI use, but opting out may reduce how well Gemini and AI Overviews can represent your brand, so it is a deliberate visibility trade-off.