Glossary

What Is Google-Extended? Google's AI Training Bot Explained

Published: 2026-03-224 min readv1.0

Google-Extended is a user-agent token introduced by Google that allows website owners to control whether their content is used to train Google's AI models, including Gemini. It is a training control mechanism -- blocking Google-Extended prevents your content from being included in AI training datasets, but does not affect your visibility in Google Search, AI Overviews, or AI Mode.

Why It Matters

Google-Extended sits at the intersection of content control and AI visibility. It was Google's answer to the growing demand from publishers who wanted to opt out of AI training without losing their Google Search presence.

Why Google-Extended matters for website owners:

  • Separate training from search. Google-Extended was specifically designed to let you control AI training independently of search. This is a significant distinction: you can block your content from training Gemini while maintaining full Google Search rankings and even appearing in AI Overviews.
  • No search impact. Unlike blocking Googlebot (which would remove your site from Google Search entirely), blocking Google-Extended has zero effect on your search rankings, indexing, or crawl budget. Google has explicitly confirmed this.
  • AI Overviews are unaffected. A common misconception is that blocking Google-Extended prevents your content from appearing in AI Overviews and AI Mode. It does not. These features are considered part of Google Search and are powered by Googlebot.
  • Industry-wide precedent. Google-Extended helped establish the pattern of AI companies providing separate user-agents for training vs. search, following OpenAI's GPTBot/OAI-SearchBot split.

Google introduced Google-Extended in September 2023 as a robots.txt control alongside updates to its terms of service regarding AI training.

How It Works

Google-Extended is not a traditional web crawler in the way Googlebot is. It functions as a policy signal within the robots.txt framework.

  • User-agent token. The token is Google-Extended. Add it to your robots.txt to create specific rules for AI training access.
  • What it controls. Blocking Google-Extended tells Google not to use content crawled from your site for training AI models like Gemini, Imagen, and other machine learning systems. Google uses content from Google Search's existing index -- it does not send a separate crawler.
  • What it does not control. Google-Extended has no effect on: Google Search indexing, Google Search rankings, AI Overviews (formerly SGE), AI Mode, Google Discover, or any other Google Search product. These remain under Googlebot's control.
  • Robots.txt syntax. Configuration is standard: User-agent: Google-Extended followed by Disallow: / to block, or Allow: / to permit. You can also control access at the directory level.
  • No separate crawler. Unlike GPTBot or PerplexityBot, Google-Extended is not a distinct crawler that makes HTTP requests to your server. It is a user-agent token that Google checks against your robots.txt policy when deciding whether to include your already-crawled content in training datasets.

Google-Extended vs. other Google user-agents

| User-agent | Controls | Blocking impact | |---|---|---| | Googlebot | Google Search indexing | Removes from Google Search entirely | | Google-Extended | AI model training | Prevents content from training Gemini | | Googlebot-Image | Image search indexing | Removes images from Google Images |

For a complete list of AI crawlers and their user-agents, see our AI crawler bots list for 2026.

Key Points

  • Google-Extended (user-agent: Google-Extended) controls whether Google uses your content to train AI models like Gemini
  • Blocking Google-Extended does not affect Google Search rankings, indexing, AI Overviews, or AI Mode
  • It is not a separate crawler -- it is a policy token checked against your robots.txt
  • You can safely block Google-Extended to opt out of training while maintaining full Google Search presence
  • Configure it in your robots.txt file with standard Allow/Disallow directives
  • See the full list of AI bots in our AI crawler bots list for 2026

Frequently Asked Questions

Does blocking Google-Extended affect my Google Search rankings?

No. Google has explicitly confirmed that Google-Extended is separate from Googlebot. Blocking Google-Extended only prevents your content from being used to train AI models like Gemini. Your Google Search rankings, indexing, and crawling remain completely unaffected. For full configuration details, see our robots.txt for AI crawlers guide.

Does blocking Google-Extended affect AI Overviews and AI Mode?

No. AI Overviews and AI Mode are considered part of Google Search. They pull content using Googlebot, not Google-Extended. Even if you block Google-Extended, your content can still appear in AI Overviews, AI Mode, and all other Google Search features. Only your inclusion in AI training datasets is affected.

What is the difference between Google-Extended and Googlebot?

Googlebot crawls the web to build Google's search index -- it powers Search, Discover, News, and all search products. Google-Extended is a separate user-agent token that only controls AI training data usage. Blocking Googlebot removes your site from Google Search entirely. Blocking Google-Extended only prevents training. For a full list of AI crawlers and their roles, see our AI crawler bots list.

How are AI crawlers treating your site?

Scan your robots.txt and see which AI bots are allowed, blocked, or misconfigured.

Scan My AI Crawler Access

Free scan -- no signup required

Google-ExtendedGoogle AI trainingGemini training botrobots.txt Google-ExtendedAI crawler