On March 20, Google added Google-Agent to its user-triggered fetchers documentation.
The documentation update is straightforward: Google added the user agent name (Google-Agent), published the associated IP ranges, and noted that Google-Agent is used by agents hosted on Google infrastructure to navigate the web and take actions on behalf of users.
Google named Project Mariner as an example in this updated documentation. Project Mariner is a research prototype that acts as an AI agent within Chrome that can complete tasks for you — access is currently quite limited.
Google noted that Google-Agent will be rolling out over the next few weeks.
How Google-Agent is different than other user agents
Google-Agent is different because it’s user-triggered. It reflects that a real person asked a Google AI agent to do something on their behalf — and the agent went to your site to do it.
Most crawlers you see in your server logs are performing background processes like Googlebot crawling your pages.
Why this matters now
Google-Agent appearing in your logs doesn’t mean AI agents are out there completing purchases and filling out forms at scale today. The protocols, standards, and functionality that will make that seamless are still being built.
What is happening is that agents are beginning to interact with the web — browsing, evaluating, and navigating content on behalf of users. That behavior is real and growing. The infrastructure around it will catch up.
That’s exactly why now is the right time to pay attention.
What to do now
Start tracking Google-Agent activity
Filter your server logs to look for Google-Agent.
Just know that volume will be low — the rollout only began March 20. That’s fine. A baseline now is what gives you context later.
Further reading: https://www.semrush.com/blog/log-file-analysis/
Check your blocking rules
Content delivery network (CDN) and web application firewall (WAF) configurations built to stop malicious bots can inadvertently block legitimate AI agents.
Semrush Enterprise AIO’s Agent Analytics provides a targeted view of which AI crawlers can access your website and how they interact with content.
And if you’re at a small- to medium-sized business, Semrush’s Site Audit tool similarly shows whether AI crawlers can effectively access your site.
Addressing any issues you see with AI crawlers’ ability to access your site now sets the stage for enabling AI agents later on. Also, ensure that the Google-Agent IP ranges published in user-triggered-agents.json are allowed for your website.
The bigger picture: Mapping the agentic future
Google is signaling we’re moving toward a web that increasingly runs on agents acting on behalf of users.
Agentic search optimization (ASO) builds on the same foundation SEO has always required but adds legibility for machines evaluating your brand on someone else’s behalf.
Understanding where the web is heading — including emerging standards like WebMCP — is crucial for staying ahead of that curve.
