Google has published a help page on how website owners can let Google-Agent traffic through their security tools. If your website sits behind Cloudflare, a hosting firewall or a security plugin, here is what Google-Agent is, what Google’s page says, and what to ask the person who manages your website.
Key takeaways
- Google-Agent is the name Google uses for AI agents running on Google’s infrastructure that visit websites and take actions when a person asks them to.
- Google’s help page explains how to allow these visits through a CDN, a web application firewall or a bot-detection service while keeping your existing security in place.
- Google says each request carries Google-Agent in its User-Agent header and comes from an IP address on a list Google publishes.
- Google’s page does not say Google-Agent affects search rankings. It is about agents acting for a user, not about crawling for Search.
First, three quick definitions
User agent: the name a browser or bot gives your website when it asks for a page.
CDN: a content delivery network, a service that sits in front of your website to speed it up and often to filter traffic.
Firewall or bot filter: a security layer (from your host, your CDN or a plugin) that blocks or challenges visitors it thinks are automated, for example with a CAPTCHA.
What Google-Agent is
Google describes Google-Agent as the name used by agents hosted on Google infrastructure to “navigate the web and perform actions upon user request.” In plain terms, a person asks a Google AI agent to do something, and the agent visits websites to get it done.
Google lists Google-Agent among its user-triggered fetchers: tools that fetch pages because a user asked, not because Google is crawling the web. That same list says user-triggered fetchers generally ignore robots.txt rules, because a user requested the fetch. So the setting that matters here is usually your firewall or bot filter, not your robots.txt file.
Google’s page does not say which Google products use Google-Agent, and it does not connect Google-Agent to search rankings.
Does this affect your business?
It may, if your site uses a security layer that challenges or blocks automated visitors. Google’s page lists the kinds of steps an agent may take on a site: on-site search, filtering, opening detail pages, add-to-cart and booking, plus the behind-the-scenes requests those features make. For a local service business, the parts that matter most are usually your booking tool, your contact or quote form and your service pages.
Google’s troubleshooting list also names common causes of blocks: broad blocks on cloud or data center networks, managed bot rules, and policies that block a whole category of AI agents. Google says these can override a verified bot setting.
Google’s page is written for sites that run one of these security tools. If your site has none, there may be nothing to change. Either way, this is technical housekeeping that sits next to the basics in our small business SEO guide, not a ranking factor.
How Google says to recognize Google-Agent
Google gives three checks:
- The name: the User-Agent header contains Google-Agent. Google’s list of user-triggered fetchers shows full mobile and desktop user agent strings, with the Chrome version left as a placeholder.
- The address: the request comes from an IP address in Google’s user-triggered-agents.json file, published on Google’s developer site.
- A signature, on some requests: Google is experimenting with Web Bot Auth, a protocol that lets a bot cryptographically sign its requests, using the identity agent.bot.goog. Google says it does not sign every request, so this is an extra check, not a replacement.
Google also cautions that a user agent string can be faked, which is why the name alone is not enough.
What Google says to do in your firewall or bot filter
Google says some bot-detection services, CDNs and firewalls already offer a way to verify Google-Agent requests and allow them, and it points you to your provider for the exact steps. If your service does not verify them automatically, Google lists these rules:
- Check the name. Confirm the User-Agent header contains Google-Agent.
- Check the address. Confirm the IP address is in user-triggered-agents.json.
- Check the signature if there is one. Verify the signature headers the way Google’s Web Bot Auth page describes.
- Skip bot challenges for requests that pass. Everything else your security tool blocks stays blocked.
If Google-Agent is still blocked
Google’s troubleshooting section lists six things to check. In plain words:
- Rule order: the allow rule must run before broad data center blocks, managed bot rules or AI agent category blocks.
- Challenges: the rule should skip CAPTCHAs and JavaScript challenges too, including CAPTCHA boxes that pop up on search or form submission.
- The whole visit: allowing the homepage is not enough. Booking, search and the background requests behind them need the same treatment.
- Rate limits: requests from many users share Google’s outgoing addresses, so per-IP rate limits should account for verified Google-Agent ranges.
- Updated ranges: IP-based rules should stay in sync with the latest version of Google’s file.
- Proxies: load balancers and proxies should pass along the original visitor IP, the User-Agent header and any signature headers.
What to ask whoever manages your website
- Do we have a CDN, firewall, bot filter or security plugin in front of our site, and which one?
- Does it already verify and allow Google-Agent, or do we need a rule for it?
- If we add a rule, does it check both the User-Agent and Google’s user-triggered-agents.json IP list, and skip challenges only for requests that pass?
- Does the rule cover our booking tool and contact forms, not just the homepage?
- How do we keep the IP list current?
Frequently asked questions
What is Google-Agent?
Google says it is the name used by agents hosted on Google infrastructure that browse the web and take actions when a user asks them to. Google lists it among its user-triggered fetchers.
Does Google-Agent affect my search rankings?
Google’s page does not say so. It is about letting agents that act for a user reach your site, not about crawling for Search.
How do I know a visit really comes from Google-Agent?
Google says to confirm the User-Agent header contains Google-Agent and the IP address is in its user-triggered-agents.json file. Some requests are also signed through Web Bot Auth.
Should I turn off my firewall or bot protection?
No. Google’s advice is to verify Google-Agent requests and skip bot challenges for those verified requests only, while keeping your other security controls in place.
Related reading: our note on the Google tag becoming a Tag Manager container, another Google change to check with your web person this month. Sources: Google’s help page on allowing Google-Agent traffic (last updated November 17, 2026, per the page), its list of user-triggered fetchers and its page on Web Bot Auth, all read on November 18, 2026.
Not sure whether your firewall or security plugin lets Google-Agent through? Bright Marketing Solutions, with offices in Brooklyn, NY and Dumont, NJ, can review your site’s setup with your web person.