The best way to track AI traffic in Google Analytics 4 changed in May 2026. GA4 now has a native AI Assistants channel for recognised referral visits from tools such as ChatGPT, Gemini and Claude.

It gives marketers a cleaner starting point than the regex-first workarounds older guides recommend.

But GA4 still does not show every AI-influenced visit. Google AI Overviews and AI Mode remain part of Organic Search, while AI crawler activity needs to be measured outside GA4.

TL;DR

• GA4 introduced a native AI Assistants channel in May 2026, so this should now be the first place to check recognised AI referral traffic
  
  

• Use Session source / medium, Source Group, landing pages, engagement, key events, and revenue or leads to judge the quality of AI-driven visits
  
  

• Google AI Overviews and AI Mode remain under Organic Search in GA4, while Search Console can provide additional generative AI visibility where its dedicated report is available
  
  

• Use a custom AI channel group only when you need broader or organisation-specific classification, rather than as the default setup
  
  

• Keep human AI referrals and AI crawler activity separate: GA4 is for user sessions, while server logs, hosting logs, or CDN tools are better for monitoring crawler activity

Start With GA4’s Native AI Assistants Channel

Since May 2026, GA4 can automatically classify recognised AI-assistant referrals into a dedicated acquisition channel.

Where to find AI traffic

Go to:

Reports → Acquisition → Traffic acquisition

Then check Session default channel group for AI Assistants.

Google announced the change on 13 May 2026 in its Google Analytics release notes. When a referrer matches a recognised AI assistant, GA4 can assign the ai-assistant medium and (ai-assistant) campaign value.

Google’s current channel documentation gives ChatGPT, Gemini, DeepSeek, Copilot and Grok as examples, while the original launch announcement also named Claude.

If you want to know how to track ChatGPT traffic in GA4, this native channel should now be your first check. The same applies when analysing AI referral traffic in GA4 more broadly.

If the expected AI traffic is missing from your reports, [No data available in Google Analytics 4? Here are the solutions] covers the most common tracking and configuration checks. 

Break AI Traffic Down by Source, Source Group, and Landing Page

The channel total tells you AI referrals exist, but source and page-level reporting tells you whether they are useful.

Compare source and Source Group

Start with Session source / medium to see the individual referrer recorded for a session.

Google added Source Group on 11 June 2026 to consolidate fragmented source values into cleaner platform groups. The update includes built-in grouping for emerging sources such as ChatGPT under OpenAI and Perplexity, and Google says Source Group is populated retroactively.

That makes it useful for tracking LLM referral traffic historically without relying only on raw source values.

Check landing pages and outcomes

To report AI traffic in GA4 properly, compare:

• Sessions
  
  

• Session source / medium
  
  

• Source Group where useful
  
  

• Landing pages
  
  

• Engaged sessions or engagement rate
  
  

• Key events
  
  

• Leads or revenue, where relevant

A source sending 100 sessions with no meaningful actions may matter less than one sending 20 visits that regularly become leads.

If you want to connect AI-driven visits with meaningful actions, our guide to [GA4 Events Tracking: How to Set Up and Measure Conversions] can help you measure the events and key events that matter. 

Understand What GA4’s AI Channel Does Not Capture

The native channel improves AI reporting, but it is not a complete measure of AI-driven discovery.

Google AI Overviews and AI Mode stay in Organic Search

Google’s default channel definitions state that AI Assistants exclude Google AI Overviews and AI Mode. In GA4, visits from those Google experiences remain part of Organic Search.

Google introduced a dedicated Generative AI performance report in Search Console in June 2026 for properties where the report is available. However, the dedicated view currently reports impressions only. 

You can break those impressions down by page, country, device and date, but it does not provide clicks, CTR or query-level reporting specifically for AI Overviews and AI Mode.

Clicks and other search performance data remain blended into the broader Web Search performance reporting, so Search Console cannot yet give you a clean count of clicks generated specifically by Google’s generative AI experiences.

So the practical measurement setup is:

• GA4 AI Assistants for recognised third-party AI referral sessions
  
  

• Search Console’s Generative AI report for impressions from AI Overviews and AI Mode where available
  
  

• The broader Search Console Web report for clicks, CTR and query performance that remains blended with other Google Search activity
  
  

• Organic Search in GA4 for visits arriving from Google AI search experiences

Referral information can still be missing

GA4 can only classify a visit using the source information available when someone reaches the site.

If useful referral context is missing, a visit may not appear under AI Assistants. Treat the channel as measurable AI referral traffic rather than an exact count of everyone who discovered your brand through an AI answer.

When to Create a Custom AI Channel Group

The native channel should be your starting point, but custom grouping still helps when you need more control over classification.

Use custom grouping as a fallback

A custom AI channel group in GA4 can be useful when:

• A source is not classified the way you expect
  
  

• You want to include a broader list of AI referrers
  
  

• Your organisation uses its own classification rules
  
  

• You need consistent historical analysis based on your chosen rules

Google’s custom channel group guidance still provides regex-based examples for grouping AI-assistant sources such as ChatGPT, Gemini, Copilot, Claude and Perplexity.

Custom channel groups can also be applied retroactively in reports, which makes them useful when you want to analyse historical referral data using your own classification logic.

Older tutorials often start with this setup. In 2026, it makes more sense to check the native AI Assistants channel and Source Group first, then add custom rules when you have a specific reporting gap.

Before adding more reporting logic, it is also worth checking the wider setup against [10 Google Analytics 4 Mistakes in Configuration You Should Avoid] so classification problems are not being caused by an underlying GA4 issue. 

AI Referral Traffic and AI Crawler Traffic Are Different

Human visits from AI assistants and automated crawler requests represent different behaviours, so they should not be reported together.

Human referral traffic belongs in GA4

If someone clicks a link in ChatGPT, Gemini, or another recognised assistant and reaches your website, GA4 may classify that session under AI Assistants when the referrer is recognised.

That is a human acquisition session and belongs in your GA4 reporting.

AI crawlers belong in server or CDN logs

AI crawlers request pages automatically so AI systems can search, retrieve, train on, or otherwise process content.

GA4 is not the right primary tool for monitoring this activity because known bots and spiders are automatically excluded from Analytics reporting.

For crawler visibility, use:

• Server access logs
  
  

• Hosting logs
  
  

• CDN analytics
  
  

• Tools such as Cloudflare AI Crawl Control

A useful rule is:

GA4 = human AI referral traffic

Server/CDN logs = AI crawler activity

What to Include in an AI Traffic Report

A useful AI traffic report should show whether AI-assisted discovery contributes meaningful visits and outcomes, not simply whether the channel exists.

Include:

• AI Assistants sessions
  
  

• Session source / medium
  
  

• Source Group
  
  

• Landing pages
  
  

• Engagement rate or engaged sessions
  
  

• Key events
  
  

• Leads or revenue, where relevant
  
  

• Comparison with Organic Search and Referral traffic

Use GA4’s AI-Powered Insights as a Reporting Layer

Google Analytics introduced AI-powered tools in August 2026 to help marketers analyse performance shifts and cut down on manual data review.

GA4 can surface AI Overviews and automated insights that summarise notable changes, anomalies and trends in performance and direct users towards the relevant reports.

Google is also rolling out Ask Advisor in beta for eligible Analytics properties. It allows marketers to ask natural-language questions about their GA4 data, request visualisations, investigate why metrics changed and explore performance against previous periods or relevant benchmarks.

These tools do not change how AI referral traffic is attributed. They provide an additional analysis layer after GA4 has collected and classified the traffic.

Review these metrics over time rather than reacting to a large percentage increase from a very small traffic base.

Together, the native channel, Source Group, Search Console, and crawler logs give you a more complete way to track AI traffic in Google Analytics 4 without assuming GA4 can measure every form of AI visibility on its own.

You can take the analysis further with [GA4 Segments and Audiences: How to Slice Your Data for Better Insights] to compare AI visitors with other traffic and audience groups. 

Need Help Building an AI Traffic Report?

We can help bring clarity if your AI traffic data is fragmented across native channels, source dimensions, Search Console, and custom reports.

At No Fluff, we can audit your GA4 setup, build a cleaner AI traffic view, and connect referral visits with landing pages, key events and business outcomes.

If you want us to review how AI traffic is being measured across your analytics setup, get in touch with us.