The Correct Way to Filter & View AI Traffic in GA4
https://www.linkedin.com/company/rem-web-solutions/
https://www.facebook.com/remwebsolutions
https://www.instagram.com/remwebsolutions/
If you searched "how to track AI traffic in GA4" any time before the middle of May 2026, the instructions you found are already out of date.
Google quietly rebuilt how Google Analytics 4 handles this traffic, and most of what's published online, including some of our own earlier advice, was written for the old system.
This guide covers what changed, what the new setup still misses, and exactly how to filter AI traffic in GA4 so that the number you report is one you can stand behind.
TL;DR
|
What Actually Changed in GA4 on May 13, 2026?
Before this update, Google Analytics 4 treated traffic from AI tools like ChatGPT and Gemini as regular referral traffic. This made it difficult to see how much of a website’s traffic was coming specifically from AI assistants.
The workaround was to create a custom channel group and manually add AI referral sources. As new AI tools appeared, you also had to keep updating those rules.
That changed on May 13, 2026.
Google confirmed the update in its GA4 release notes. Search Engine Journal also reported the change.
GA4 now has a dedicated AI Assistant channel in its Default Channel Group. When GA4 recognizes traffic from a supported AI assistant, it automatically:
Sets the medium to ai-assistant
Groups the session under the AI Assistant channel
Sets the campaign to (ai-assistant)
This means you no longer need to create a custom channel just to identify this type of traffic.
The feature rolled out gradually, with reports of it appearing across more properties through late May and early June 2026. So, if you don't see AI Assistant traffic in your reports yet, it may simply be part of the rollout rather than an issue with your GA4 setup.
What Does the Native AI Assistant Channel Actually Track?
The new AI Assistant channel is useful, but there’s an important detail to understand. It does not capture every type of AI-driven traffic.
When Google first announced the update, it mentioned ChatGPT, Gemini, and Claude as examples. However, Google's current GA4 documentation lists five AI assistants:
Did you notice that Claude and Perplexity are not on the current list?
Traffic from these platforms can still appear in GA4, but it will generally be classified as Referral unless you create a custom channel to track them separately.
There is another important distinction. Google's own AI Overviews and AI Mode are not included in the AI Assistant channel.
Clicks from these Google Search features are still counted as Organic Search, because users are coming from Google's search results rather than from a separate AI assistant website.
So, when looking at AI traffic in GA4, think about it in two groups:
Third-party AI assistants such as ChatGPT, Gemini, DeepSeek, Copilot, and Grok. These can appear on the AI Assistant channel.
Google AI features such as AI Overviews and AI Mode. These remain part of the Organic Search.
Google defines the AI Assistant channel as traffic coming from sources such as ChatGPT, Gemini, DeepSeek, Copilot, and Grok, while specifically excluding AI Overviews and AI Mode.
The key takeaway: Don't treat the AI Assistant channel as your total AI traffic. It gives you a useful starting point, but it is more of a floor than a ceiling. If you want a complete picture of how AI is driving visitors to your website, you'll need to look beyond this one GA4 channel.
How to View AI Traffic Using the Native Channel (Step by Step)
Checking your AI traffic in GA4 is now simple and only takes a minute.
Log in to your GA4 property and open Reports.
Go to Acquisition > Traffic acquisition.
Make sure the primary dimension is set to Session default channel group.
Scroll through the table and look for AI Assistant, alongside channels such as Organic Search, Direct, Paid Search, and Referral.
To see where the traffic came from, change the dimension to Session source/medium. Look for sources with the medium ai-assistant.
One Important Limitation: Keep in mind that the GA4 AI Assistant update is not retroactive.
This means GA4 only applies the new classification to traffic going forward. Traffic from before the rollout won't be moved into the AI Assistant channel later. Older AI-driven visits may still be sitting under Referral, Direct, or another channel.
So, if you're comparing AI traffic before and after May 2026, be careful with the numbers. A lower AI Assistant number for earlier periods doesn't necessarily mean your website received less AI traffic. Some of that traffic simply wasn't classified as AI traffic at the time.
Why the Native Channel Isn't Enough on Its Own
The native AI Assistant channel makes AI traffic much easier to track, but it still doesn't give you the full picture.
There are three important gaps to keep in mind.
The Named-Platform Gap
Google's current list includes platforms such as ChatGPT, Gemini, DeepSeek, Copilot, and Grok. Perplexity and Claude are not currently included. If these platforms send traffic to your website, GA4 may classify those visits as regular Referral traffic instead of AI Assistant traffic. That means your AI traffic numbers could be lower than they are.
The Google AI Gap
Traffic from Google AI Overviews and AI Mode is still counted as Organic Search. This is important because these Google AI features can appear directly within search results. GA4 currently doesn't provide a separate native channel that lets you see exactly how many clicks came from these AI experiences. So, some of your AI-driven search traffic is effectively hidden inside your regular organic traffic numbers.
The No-Referrer Gap
Some AI-referred visits may arrive without usable referral information, particularly from apps or in-app browsers. When GA4 can't identify the source, it may classify the visit as Direct. This means some AI-driven visits can be hidden within your Direct traffic and cannot be reliably identified in GA4 alone.
The bottom line: The native AI Assistant channel is a helpful starting point, but it shouldn't be treated as your complete AI traffic report. To understand the full impact of AI on your website, you'll need to look at AI Assistant, Referral, Organic Search, and Direct traffic together.
How to Build a Custom Channel Group That Catches Everything
A custom channel group remains the most reliable way to see the full picture, because it's stable, it isn't limited to Google's named list, and it survives quiet changes on the vendor's side. Google Analytics 4 allows a maximum of 2 custom channel groups per property for standard (free) accounts, and up to 5 custom channel groups for GA4 360 properties. Each custom channel group can contain up to 25 individual channels, giving you plenty of room to build out advanced attribution.
Go to Admin > Data display > Channel groups.
Click Create a new channel group and give it a clear name such as "AI Traffic (All Sources)."
Add a new channel to that group. Set the condition to Source matches regex, and include the platforms relevant to your audience, for example:
chatgpt|openai|gemini|claude|anthropic|perplexity|copilot|deepseek|grok|meta\.ai|you\.com.
Defining this at the channel group level permanently groups these sources into a single clean dimension, preventing your AI traffic data from staying fragmented across dozens of individual rows in your standard reports.
Order matters. GA4 evaluates channel rules from top to bottom, and the first match wins. If your AI channel sits below Referral in the list, every AI session gets claimed by the more generic Referral rule first, and your custom channel will show nothing. Move it above Referral and save.
Review the data under Session custom channel group in your Traffic acquisition report or open it as an Exploration for deeper segmentation by landing page and conversion.
Running the native channel and a custom channel group side by side gives you both convenience and coverage. The native channel needs no maintenance, and the custom group is your insurance policy against Google's list changing without notice.
Why Is My AI Traffic Showing Direct?
This is one of the most frequent questions businesses have when they start looking at AI traffic in GA4.
Sometimes, a visitor discovers your website through an AI assistant, but GA4 records the visit as Direct traffic. Why? Some AI mobile apps and in-app browsers can remove the referral information before sending the visitor to your website.
When that happens, GA4 doesn't have a source to identify. It simply records the session as Direct, just like someone who typed your website address directly into their browser.
This type of traffic is sometimes described as dark traffic or unattributed AI traffic. Unfortunately, there is no perfect way to identify or measure it using GA4 alone.
However, a couple of checks can help you spot possible AI-driven visits:
Look for unusual spikes. If Direct traffic to a specific page increases shortly after AI activity or mentions related to that page, it could be a sign that AI is contributing to the increase.
Check Google Search Console for additional AI search visibility data. Google began rolling out dedicated reporting for AI Overviews and AI Mode in June 2026, helping eligible websites understand how they perform in Google's generative AI search experiences. This complements GA4, where these visits remain part of Organic Search.
Neither method will give you an exact number. But together, they can help you make a more informed estimate of how AI may be influencing your traffic.
For a deeper look at the traditional way of identifying AI referral traffic, see our companion guide: Is ChatGPT Sending Traffic to Your Website? How to Track It. It explains how AI bot activity and actual human clicks from AI citations are different and how you can track them.
Why This Actually Matters for Your Business
This isn't just about better GA4 reporting. AI traffic is becoming an increasingly important source of qualified website visitors.
According to Adobe Digital Insights, AI-referred traffic to U.S. retail websites grew 269% year over year in March 2026, while AI-referred visitors converted 42% better than non-AI traffic.
Why? Visitors coming from AI assistants may already have researched your business, compared options, or narrowed their choices before clicking through.
If that traffic is being counted as Referral, Organic Search, or Direct, you may be underestimating the value of AI in your marketing performance.
Understanding where AI traffic comes from helps you make better decisions about your SEO, content, and overall marketing budget.
For a broader look at how different channels work together, see our guide on Why SEO and Google Ads Work Better Together. And, if you're still building your understanding of how AI tools decide what to cite in the first place, How to Make AI Cite Your Website as a Primary Source is the right next read.
Closing Thoughts
GA4's new AI Assistant channel is a welcome improvement. It gives businesses a clearer way to track traffic coming from major AI platforms instead of grouping it all under Referral.
But it's not a complete picture.
For a better understanding of AI-driven traffic, businesses should combine the native AI Assistant channel with custom tracking, monitor Direct traffic for possible AI patterns, and understand which AI sources GA4 does and doesn't include.
If you're unsure whether your GA4 setup is capturing AI traffic properly, our team at REM Web Solutions can help you review your analytics and reporting. Call us now!
Stay updated on Google's algorithm changes, Canadian copyright insights, and web design best practices. Subscribe to the REM Web Solutions monthly newsletter for practical strategies and industry insights delivered straight to your inbox.
Frequently Asked Questions
