Google-Agent · User agent of Google’s AI agents
Google-Agent is the user agent used by AI agents hosted on Google infrastructure to browse the web and perform actions on a user’s request, for example Project Mariner. Google lists it among user-triggered fetchers, which generally ignore robots.txt.
Standalone definition · 38 words
When Google added Google-Agent to its documentation
Google added Google-Agent to its list of user-triggered fetchers on March 20, 2026, with a rollout announced over several weeks. The documentation names Project Mariner, Google’s experimental agent that completes tasks in a browser.
How to recognise Google-Agent in your logs
The string contains compatible; Google-Agent, in mobile and desktop versions. IP addresses come from the user-triggered-agents.json file published by Google. Google is also testing the Web Bot Auth protocol with the agent.bot.goog identity to authenticate these requests.
Can you block Google-Agent with robots.txt?
Not reliably. Google says user-triggered fetchers generally ignore robots.txt rules, since a person requested the visit. The real topic is making the site usable by an agent, see agent-friendly website.
Frequently asked questions about Google-Agent
Does Google-Agent crawl for Google Search?
No. It acts for a specific user. Indexing for Search remains Googlebot’s job.
Should you worry about fake Google-Agent traffic?
As with Googlebot, verify the IP against Google’s published files before trusting a user agent.
Terms related to Google-Agent
Further reading on Google-Agent
- Google’s GEO checklist: how to optimize your site for AI agents
- Search Central Live Bengaluru 2026: what Google said
- ← Back to the GEO glossary (126 terms)