Google Analytics counts a visit only when a snippet of JavaScript runs in the visitor's browser and reports back. That design is the whole reason AI bots are missing from your dashboard: when an assistant sends an agent to read a page, the agent requests the raw HTML, takes the text, and leaves without ever executing that snippet. No script, no event, no record. The page was read and the content was used, but as far as your analytics are concerned nobody came, which is how a large and growing share of real traffic ends up completely invisible.
How Google Analytics actually counts a visit
Google Analytics, like almost all client-side analytics, works by running a small piece of JavaScript in the browser. When a human loads a page, the browser downloads the HTML, then executes the scripts on the page. The analytics tag fires during that step, sending a pageview event back to Google with details about the visit. No JavaScript execution means no event. The entire model assumes the visitor is a browser that runs code.
Why AI bots break that model
AI bots are not browsers and they do not behave like one. They make an HTTP request, receive the HTML, and read it as text. There is no rendering step and no script execution. This is true of training crawlers that gather content in bulk, and it is especially true of Live Search Agents, the bots dispatched in real time to answer a specific user question. None of them trigger the analytics tag, so none of them appear in your dashboard. The same gap hides them from your ad stack, your affiliate links, and your paywall logic, all of which also live in JavaScript.
What this blind spot costs you
The consequence is that publishers make editorial and commercial decisions on incomplete data. If a third of your real traffic is AI agents and none of it appears in analytics, then your most-read content, your true audience size, and your most valuable emerging segment are all mismeasured. You may be deprioritising content that AI systems retrieve heavily because your dashboard says it is quiet. You are also leaving that traffic unmonetised, because you cannot monetise, or even value, an audience you cannot see.
How to actually see AI bot traffic
To measure AI bots you have to move below the browser, to the server or the CDN edge, where the raw request is visible before any rendering. Server log analysis is the traditional route: parse your access logs and classify requests by user agent and source. The limitation is that user-agent strings can be spoofed and raw logs are noisy and hard to attribute. A more reliable approach is to classify traffic at the CDN edge, where you can verify a bot against its owner's published IP ranges and signatures and type it accurately in real time. blankspace does this as its analytics layer, capturing every AI agent, typing it as a Live Search Agent, training crawler, search crawler, or scraper, and attributing it to its owner, all before JavaScript would have run. The analytics are free and go live on the first request.
What to look for once you can see it
Once AI traffic is visible, the useful questions are which agents visit most, which pages they retrieve, and how the volume is trending. In live publisher accounts, Live Search is frequently the single largest category of AI traffic. Seeing that mix changes content strategy, because the pages AI systems retrieve most are often not the pages your human analytics flagged as important. It also reveals the monetisation opportunity, since each Live Search retrieval is a high-intent read happening at the moment a user is researching a decision.
Frequently asked questions
Do any AI bots run JavaScript?
A small number of agents can render JavaScript in some modes, but the large majority of training crawlers and Live Search Agents fetch and parse raw HTML without executing scripts. You cannot rely on client-side analytics to capture them.
Can I see AI bots in Google Analytics with a filter or setting?
No. The issue is not a filter. The events are never sent, because the analytics JavaScript never runs. No configuration in Google Analytics can recover a visit that did not fire the tag.
Will server logs show me all my AI traffic?
Server logs show the requests, but attributing them accurately is hard because user-agent strings can be spoofed and logs are noisy. Verifying bots against owner IP ranges and signatures, typically at the CDN edge, gives a far more reliable picture.
How much of my traffic could be invisible?
It varies by site, but publishers commonly find a large and growing share of real traffic is AI agents that never appear in client-side analytics. In some accounts AI agents are the single largest traffic category.
Does seeing AI traffic let me monetise it?
Seeing it is the prerequisite. Once AI agent retrievals are visible and classified at the edge, eligible Live Search retrievals can be monetised through content-layer advertising, which is the only mechanism that reaches bots that do not run JavaScript.