How to track ChatGPT and AI traffic in GA4

You cannot improve what you cannot see. GA4 now has a built-in AI Assistant channel, which is the place to start, but it does not show everything: some AI clicks arrive with no referrer, and any that also carry no source tag can end up in 'Direct.' Here is how to read the native channel, add a custom one covering ChatGPT, Perplexity, Claude, Gemini and Copilot, and why even then the number is a floor.

Marketers keep asking "how much traffic are we getting from AI?" and then trusting whatever GA4 shows. The honest answer is: probably more than GA4 shows, and reading even that much correctly takes a little care. Both problems are workable. For the market picture, which assistants send clicks at all and how unevenly, see what 6.77 million sessions say about ChatGPT and AI referral traffic; this guide is about seeing your own number.

Step 1: start with GA4's built-in AI Assistant channel

On May 13, 2026, Google added an AI Assistant channel to GA4's Default Channel Group, along with an ai-assistant medium and an (ai-assistant) campaign value. Google's channel documentation describes it as the channel by which users arrive from sources like ChatGPT, Gemini, Deepseek, Copilot or Grok, and the release note also names Claude. When a visit's referrer matches Google's list of AI assistants, GA4 sets the medium to ai-assistant and the session lands in AI Assistant with no setup on your part.

Two limits are worth knowing before you rely on it. Google says the channel excludes its own AI Overviews and AI Mode, so clicks from those surfaces do not appear in it. And Google does not publish the full list of referrers it matches, so check where Perplexity and smaller assistants land in your own property: in AI Assistant, or still in Referral.

Step 2: add a custom AI channel group for control

The native channel needs no setup, but you cannot see or edit its list. A custom channel group lets you decide which hosts count as AI, and Google says custom channel groups can be applied to your reports retroactively, so you can trend AI traffic from before May 2026 too. To build one:

1. Go to Admin, then Channel groups, and create a new custom channel group.
2. Add a channel called AI (or "AI Assistants").
3. Set the rule to match Source matches regex, and use a pattern covering the major AI hosts:

Host-based: chatgpt\.com|perplexity\.ai|claude\.ai|gemini\.google\.com|copilot\.microsoft\.com|meta\.ai

A broader token pattern catches variants and subdomains: .*chatgpt.*|.*openai.*|.*perplexity.*|.*claude.*|.*gemini.*|.*copilot.*.

Then move the AI rule to the top of the ordering so those sessions get claimed by your AI channel before another rule grabs them.

"GA4 now shows some of your AI traffic by default. Some of the rest can end up in Direct."

Curious whether AI engines name you when buyers ask about your category? Get a free visibility audit and see where you stand across ChatGPT, Gemini, Perplexity and Google AI Overviews.

Step 3: understand the "Direct" undercount

Here is a common catch. Some AI clicks can arrive with no referring URL. In one sample, Clickport found that 35.7% of recognized AI search sessions on its customers' sites (935 of 2,619, in the 30 days to April 23, 2026) had no recorded referrer and were identified only by their source tags. Clicks that have neither a referrer nor a tag have nothing to attribute them to, and GA4 files them under Direct.

The consequence: even a perfectly built AI channel group shows you a floor, not the true total. Some of your real AI traffic can end up in Direct, where it looks the same as bookmarks and typed-in visits. Read your AI channel number as "at least this much," not "exactly this much."

What to do with the data

• Trend it, do not trust the absolute. Because of the Direct undercount, the direction and relative movement of your AI channel matter more than the exact number.
• Segment by engine. Break the AI channel into ChatGPT, Gemini, Perplexity and the rest, so you know which assistants actually send you people.
• Pair it with citation tracking. Referral traffic tells you who clicked through. It cannot tell you whether you were named in answers that never sent a click, which can be a large share of them. Measure citation directly too.

The takeaway

Three steps get you most of the way: start with the native AI Assistant channel, add a custom AI channel group with the right regex for control and history, and remember that some AI traffic can hide in Direct with no referrer. Do all three and you will see more of your AI traffic than any single report shows, while remembering that clicks are only part of the story, since much of your AI exposure can come from being named in answers that never send a click at all.

Common mistakes that quietly break the setup

A broken AI channel is often not a regex problem. Small ordering and scoping mistakes can break it without any warning in GA4. If your AI number looks suspiciously flat or large, start here before touching the pattern.

• Rule order left at the bottom. Channel rules are evaluated top to bottom, and the first match wins. If a broad "Referral" or "Organic" rule sits above your AI rule, it can claim ChatGPT sessions before your AI channel ever sees them. Your regex can be perfect and still capture almost nothing.
• Anchoring the pattern too tightly. A rule like ^chatgpt\.com$ matches only that exact string and misses any variant of the host, such as a subdomain. If you want reliability, match on a token that appears in the host rather than demanding an exact equality.
• Confusing source with medium. Sessions GA4 recognizes as AI assistants get the medium ai-assistant, while AI referrals it does not recognize typically arrive with a medium of "referral" and a source that is the AI host. Build your custom rule on source so it catches both. A rule on medium alone either misses the unrecognized assistants or lumps them in with every other referrer.
• Comparing native and custom channels over older dates. Google says a custom channel group can be applied to your reports retroactively. Its documentation does not say whether sessions from before May 13, 2026 are reclassified into the native AI Assistant channel, so check before comparing the two over older date ranges, where a gap may be expected rather than a bug.

None of these raise an error in GA4. They just hand you a plausible number that is wrong - the most expensive kind of mistake in a report your team acts on.

A worked example: reading the report honestly

Say your custom AI channel shows 1,200 sessions last month, and Direct shows 40,000. The naive read is "AI is a rounding error, ignore it." That read may well be wrong, and here is how to pressure-test it without pretending you have numbers you do not.

Start with the shape of Direct. A healthy site with steady branded search should have a fairly stable Direct channel. If Direct has climbed month over month while your obvious campaigns have not, some of that climb may be referrer-less AI traffic that GA4 could not label. You will not get an exact figure, but you can see the drift, and the drift is worth investigating.

Next, look at what those Direct sessions do. Filter Direct to landing pages that are deep and unlikely to be bookmarked - a comparison page, a pricing breakdown, a long guide. Bookmark and typed-in traffic tends to land on the homepage and a few shallow pages. A stranger landing cold on your "X versus Y" page with no referrer looks a lot more like someone who just asked an assistant which tool to pick.

"The AI number GA4 hands you is a floor. The interesting question is how far above the floor you actually are."

The point is not a precise second number. It is to stop treating the AI channel figure as the ceiling. Once you read it as a floor and check Direct for corroborating shape, you make far better calls about where AI sits in your mix.

Why clicks are the smaller half of the story

Even a flawless GA4 setup only ever measures the people who clicked. That was a reasonable proxy in the search era, when ranking well and getting the click were nearly the same event. In AI answers they often are not. The assistant reads your page, synthesizes an answer, and can resolve the user's question inside the chat. The user never arrives, so GA4 never sees them, no matter how good your channel group is.

This blind spot matters whenever a query ends in an answer instead of a link. If your brand is named, summarized and recommended inside the response, that is a real marketing outcome with zero sessions attached to it. Conversely, you can win the click war on the small slice that clicks through and still be invisible in the answers that shape the buying decision.

So treat GA4 as one instrument, not the whole dashboard. It tells you who the AI channel sent to your door. It cannot tell you whether the assistant named you, cited you, or recommended a competitor while your page sat unread in its context window. Measuring that means watching the answers themselves - visibility a referral report was never built to give you.

Where you can see citations, not just clicks

Some engines now report citations directly. Bing Webmaster Tools added an AI Performance report in public preview on February 10, 2026. It shows how your content is cited across Microsoft Copilot, AI-generated summaries in Bing and select partner integrations, with total citations, average cited pages, the grounding queries behind them, page-level citation activity and trends over time. On June 16, 2026, Bing began rolling out four more views in preview: intents, topics, citation share (the percentage of citations attributed to your site out of all citations shown across all sites for the same grounding query) and a compare view.

The report is built around citations rather than traffic, so it pairs well with GA4: Bing tells you when you were cited, GA4 tells you who clicked through. Read it for trends rather than exact counts. When the numbers jumped from June 1, 2026, Search Engine Roundtable reported that Bing's Krishna Madhavan described it as regular data backfilling, with no anomalies.

On the Google side, Search Console also has a Generative AI performance report under Performance. It is worth checking alongside the GA4 channel.