Fix GA4’s AI Assistant Channel That Undercounts Traffic

While sifting through a Google Analytics 4 property recently, I paused on a traffic breakdown by source and medium. The same source, chatgpt.com, appeared in three different channels simultaneously. Not three different sources. One source, scattered across three channels within a single report. If your GA4 setup includes the new AI Assistant channel, there is a strong likelihood the same fragmentation is silently distorting your data. This means the AI traffic figure you are currently reporting is almost certainly inaccurate.

AI traffic remains a modest portion of total site visits for most properties, but its conversion rate punches well above its weight. Similarweb’s clickstream data shows ChatGPT referrals converting at approximately 7%, a rate that surpasses organic search and rivals paid traffic. A high-intent channel this small deserves precise measurement, not a rough estimate. Let me explain why the fragmentation occurs and how to resolve it.

What Did Google Actually Change?

On May 13, 2026, Google introduced a native AI Assistant channel into GA4’s Default Channel Group. The concept is straightforward. When GA4 identifies a referrer it recognizes as an AI assistant, it assigns the session the medium `ai-assistant`, places it in the AI Assistant channel, and labels the campaign as `(ai-assistant)`. No configuration, no regex, no manual setup required. The feature rolled out gradually, reaching most properties by early June 2026.

If you spent the past year crafting custom regex just to isolate your AI traffic, this is a genuine improvement. Previously, those visits were buried in Referral or Direct when the referrer was absent.

However, the catch lies in the word “recognizes.” The list of platforms Google acknowledges is fluid. At launch, it included ChatGPT, Gemini, and Claude. By June, the live documentation listed a different set: ChatGPT, Gemini, DeepSeek, Copilot, and Grok, with Claude quietly removed. Perplexity, one of the highest-intent AI sources, remains absent from the list and continues landing in Referral. So avoid hardcoding a platform list into client reports. Check Google’s channel definitions on the day you publish, because they change.

1 Source, 3 Channels: The Problem Hiding in Your Reports

Returning to that screenshot. The reason a single source splits across three channels is that GA4 determines the channel using source and medium together, not source alone. Add `Session source/medium` as a dimension, and you can watch `chatgpt.com` fracture into three segments:

• `chatgpt.com / ai-assistant` lands in AI Assistant. This is the slice GA4 recognized and tagged.Why would a recognizable source appear with no medium? In my experience, it is usually the ChatGPT app and its in-app browser. As MarTech has documented, links opened inside those embedded browsers tend to strip the referrer, so GA4 retains the source but loses the medium.Thus, your real ChatGPT number is not in one place. It is smeared across three channels, and one of them is a bucket you probably never check. Reading the AI Assistant channel alone will undercount every single time.Why the Obvious Fixes Don’t WorkThree instinctive fixes exist, and I would not rely on any of them:
• Just read the AI Assistant channel. It misses the Referral and Unassigned slices of the same traffic, and it ignores Perplexity entirely.The Fix: 1 Custom Channel That Matches on SourceThe move that actually solves this is building a custom channel group that matches on source while ignoring medium completely. The moment you do that, the `ai-assistant`, `referral`, and `(not set)` versions of `chatgpt.com` collapse into a single line.You gain two bonuses. A custom channel group applies its rules retroactively across your entire date range, rescuing all those old ChatGPT sessions stuck in Referral. And because you write the rule, you can include the platforms Google leaves out, Perplexity included.Here is how I set it up:1. Go to Admin > Data display > Channel groups and create a new group.
  1. Add a channel and call it AI.
  2. Set the condition to Source matches regex, with a pattern covering the AI domains.
  3. Drag the AI channel above Referral and Organic, so it claims those sessions first.
  4. Save, then apply the group as your primary dimension in any acquisition report.The regex is where you can quietly sabotage yourself, so build it carefully. Keep entries tied to recognizable domains or service-specific host tokens. Never throw in a bare token like `gpt` on its own, because it will match any source containing those three letters and drag in false positives. Here is a boundary-aware pattern covering the major AI sources you would want to track:`.(^|[/.:@?&=])(chatgpt\.com|chat-gpt\.org|openai\.com|perplexity|gemini\.google\.com|copilot\.microsoft\.com|edgepilot|edgeservices|claude\.ai|deepseek\.com|grok\.com|you\.com|nimble\.ai|iask\.ai|aitastic\.app|bnngpt\.com|writesonic\.com|copy\.ai)([/.:@?&#=]|$).`Two quick notes. `gemini.google.com` is a specific host, so it will not hoover up all your Google organic, which is why you must never add a bare `google` to the list. Treat the pattern as perishable. Review it every quarter, because platforms come and go and their domains may change.The Bit Even a Perfect Channel Can’t FixThis is the part most guides skip, and it keeps you honest. A custom channel group fixes how traffic is classified, not whether it gets collected in the first place. A source rule can only catch sessions that showed up with a source it can actually read.The biggest blind spot is AI traffic with no referrer at all, which lands in Direct. Most of that comes from the AI mobile apps and in-app browsers, which pass nothing for GA4 to read, so there is no source to match on.On top of that, Google’s own AI Overviews and AI Mode get counted as Organic Search (`google / organic`), and Google deliberately keeps them out of the AI Assistant channel. For a lot of sites, that is the single biggest AI surface there is, and it is invisible as “AI” in your reports. Do not try to drag it into your AI channel either, or you will swallow up ordinary Google organic by accident.So be precise about what you are claiming. With this custom channel in place, you have a complete and consistent number for the AI sources you can actually identify. What you have not got is a measure of total AI influence. Whether an assistant recommended you to someone who never clicked is a completely different problem, and no channel group will solve it.What I’d Do This WeekTreat the native AI Assistant channel as a starting point, not the finished product. Then:
• Add Session source/medium to a Traffic acquisition report and go looking for your AI domains spread across AI Assistant, Referral, and Unassigned. Seeing the split for yourself is half the lesson.Google finally putting AI traffic on the dashboard is a real step forward. Just remember that the default view hands you the easy slice. The actual work now is knowing exactly which slices it is quietly leaving out.