AI Search Statistics: Usage, AI Overviews and Click Data

Quick answer: AI search adoption keeps climbing across Google AI Overviews, AI Mode and chat assistants. Several studies show fewer website clicks when an AI answer appears. Audience size, search volume and referral traffic measure different things, so compare each dated figure inside its own scope and judge your content by your own clicks, leads and citations.

Last updated: November 5, 2026. We update this page monthly. Every figure links to its source and shows the source date.

Google says AI Overviews reached over 2.5 billion monthly users in May 2026. An Ahrefs study of 300,000 keywords found that the top-ranking page gets a 58% lower average click-through rate when an AI Overview appears. More people see AI answers while fewer clicks leave the results page, and each platform reports usage in a different unit. For marketers, that means the audience gets bigger but the traffic report doesn’t show it.

Both of those figures were measured. Gartner predicted a 25% drop in traditional search engine volume by 2026, driven by AI chatbots and virtual agents, but that was a 2024 forecast, not a measured result. The tables below separate three things: adoption, answer visibility and website visits.

Before You Read On

• AI Overviews reach counts users. Appearance rate counts queries. They are different measures.
• ChatGPT reports weekly users, Gemini reports monthly users and Perplexity reports queries. The numbers don’t line up one to one.
• Several studies link AI answers to fewer website clicks, but each one uses a different sample.
• An AI referral is a website visit. A citation, a brand mention or an answer read without a click is not.

AI Overviews and AI Mode Reach

AI Overviews sit inside Google results. AI Mode is a separate conversational search experience. Reach figures count users, while prevalence studies count how often an answer shows up in a tracked set of queries.

No single AI Overview appearance rate holds across these datasets. Branded, commercial and broad query samples answer different marketing questions. Audit the searches your buyers actually run before you apply any industry percentage to your content plan.

ChatGPT, Gemini, Perplexity and Copilot Usage

These figures mix audience counts, query volumes and growth rates. Keep those labels intact when you compare products or brief leadership.

Assistant adoption does not directly measure search activity or buying intent. Perplexity’s query count is not a user count, and the 1 billion weekly figure is a projection, not a result. The Copilot audience count and the Perplexity disclosure date from 2025, and we show them with their original dates.

AI Search Market Share, Organic CTR and Referral Traffic

Search-like requests, organic click-through rate (CTR) and referred visits belong in separate reporting columns. The two ChatGPT search-volume estimates below show why no single AI search market share figure settles the question.

OpenAI and Ahrefs use different methods and dates, so neither settles ChatGPT’s share of search. The click studies link AI answers to fewer outbound clicks. They do not prove the same loss for every website, and we recheck the live referral tracker at each monthly update.

User Behavior and Trust

Adoption is not trust, and news questions are not every search task. Keep the Reuters Institute findings tied to their news context.

Many people use AI answers without fully trusting them. Clear evidence and accurate attribution still matter when an assistant delivers the first answer. News click behavior does not predict how buyers research software, professional services or other purchases.

How to Read These Numbers

The most useful AI search statistics keep their denominator, geography, time period and attribution. Company-reported usage is not independently audited. Studies use different samples, countries and definitions. Six problems come up most often when you apply these figures to your own reporting.

• Overlapping audiences distort market share. One person using three assistants shows up in three product totals. Adding those audiences does not give you a unique population, and dividing them by Google users does not give you search share. Weekly and monthly activity also cover different windows.
• Different samples change the rate. A branded-search dataset reflects different questions than a commercial-keyword panel. Device and geography shift the result too. Treat each series as its own indicator and compare periods within that series.
• Some AI visits arrive unlabeled. When an AI app or chat interface passes no referrer, the visit lands in analytics as direct traffic instead of an AI referral. Treat your AI referral count as a floor, not a full total.
• Citations shift between checks. An assistant can cite your page for a commercial question one week and drop it the next, after a model update or a change in prompt wording. One citation check is a snapshot, so ask the same questions on a fixed schedule and track the trend.
• Zero-click exposure is hard to value. When an AI Overview answers a high-intent query on the results page, the buyer sees your brand without creating a session you can report. Pair branded search impressions, direct traffic and a “how did you hear about us” field on lead forms to show leadership that influence.
• Old disclosures look fresher than they are. Publication dates do not always match measurement dates. Keep older figures labeled with their original dates, forecasts labeled as forecasts and projections labeled as projections.

Update Log

November 5, 2026: page published with the figures above.

What These Statistics Mean for Your Content

Use these numbers to set priorities, then test those priorities against your own performance. Our guide to how AI search is changing SEO connects these shifts to content planning.

Publish direct answers backed by evidence. Keep your source pages current, and review every claim before it goes live. Answer Engine Optimization (AEO) and Generative Engine Optimization (GEO) belong next to search visibility and conversion tracking, because more AI usage does not guarantee more referral traffic.

Check Your Own Data Before You Shift Content Budget

Industry averages give you context. Your own data should decide where the budget goes. Before you move budget, check three things:

• Search Console clicks and impressions for the queries that bring you leads.
• Analytics referrals from AI assistants, tracked as their own channel.
• Whether ChatGPT and Perplexity cite your pages when buyers ask your core questions.

AEO Agent X connects Google Search Console, Google Analytics and Ahrefs. It tracks positions, clicks, indexation, leads and AI citations for every URL. Nothing goes live until you approve it. Request access, or start with the free instant site report to see whether an AI answer engine uses your site as a source for five buyer questions.

Frequently Asked Questions (FAQs)

How often does an AI Overview appear in Google search results?

No single rate applies across the published studies. Ahrefs found AI Overviews on 82.91% of US desktop branded searches on September 29, 2026, while Semrush found them on 88% of its tracked commercial keywords in August 2026. Use the study that matches your market, device and query type.

How many people use ChatGPT, Gemini and Perplexity each month?

The companies don’t publish comparable monthly counts. ChatGPT reports weekly active users, Gemini reports app users and Perplexity reports queries. A query count is not a user count, so compare each figure only within its own label.

What is ChatGPT's search volume compared to Google Search?

OpenAI said in November 2026 that ChatGPT handles roughly one fifth of Google’s search volume, counting any prompt that triggers a web search. An Ahrefs estimate from February 2026 put it at about 12% for traditional Google-style queries. The two use different methods, so read them as a range, not a settled share.

Does an AI Overview reduce clicks to my website?

Pew’s 2025 browsing study and Ahrefs’ 2026 click-through study both link AI summaries to fewer clicks in their samples. Neither study can tell you what happens to a single page. Check your own clicks, impressions, leads and query mix before you blame AI Overviews for a decline.

What percentage of Google searches are zero-click searches in 2026?

SparkToro’s analysis of Similarweb panel data found that 68.01% of US Google searches from January to April 2026 ended without a click. The 2024 figure of 60.45% came from a different panel, so the two numbers show a direction of change, not an exact increase.

What percentage of website traffic comes from AI chatbots?

Statcounter put AI chatbots at 0.14% of worldwide referrals in October 2026. That benchmark does not describe your site’s traffic mix. Use your analytics referral data to measure the visits that actually reach you.