AI Search Statistics 2026: The Numbers That Matter

AI search statistics in 2026 tell one fast-moving story: 68% of US Google searches now end without a click to the open web (SparkToro × Similarweb), ChatGPT serves roughly 900 million people a week (OpenAI), and the visitors AI assistants do refer convert at 4.4 times the rate of traditional organic traffic (Semrush). Discovery is shifting from ranked links to synthesized answers, and the numbers below — adoption, click impact, citation behavior, referral value and crawler economics — are compiled with full sourcing in our free State of AI Search report.

How big is AI search in 2026?

Bigger than most marketing plans assume, and still compounding. OpenAI announced 800 million weekly active ChatGPT users at DevDay in October 2025 and has reported roughly 900 million weekly actives through 2026 — about double the figure from a year earlier. Perplexity, the largest of the pure answer engines, disclosed 780 million queries in May 2025 with growth above 20% month over month, and most trackers place it near a billion queries a month by mid-2026.

Google answered by making search itself generative. AI Overviews — the synthesized answer block above traditional results — reached 2 billion monthly users by mid-2025 on Google's own disclosure, the fully conversational AI Mode passed 1 billion monthly users by mid-2026, and the Gemini app roughly doubled year over year to around 900 million monthly actives. How the assistant experience and classic search now split the discovery job is the subject of our ChatGPT vs Google Search comparison.

How much of search is already AI-mediated depends on who measures, and the spread is instructive. Semrush recorded AI Overviews on roughly 16% of US queries in late 2025 after a mid-year peak near 25%; BrightEdge, which tracks commercial verticals, measured about 48% of monitored queries by early 2026. Both can be true at once: coverage concentrates on informational and consideration-stage queries — exactly the territory where brands earn their pipeline.

What do AI answers do to organic clicks?

They cut them roughly in half where they appear. Pew Research Center analyzed about 68,000 real Google searches from roughly 900 US adults in March 2025: when an AI Overview appeared, users clicked a traditional result 8% of the time versus 15% without one, links inside the Overview drew clicks on about 1% of views, and sessions ended outright after 26% of AI-answered searches versus 16% of traditional ones.

The keyword-level studies show the squeeze intensifying as the feature matures. Ahrefs' March 2025 analysis of 300,000 keywords found a 34.5% lower click-through rate for the top-ranking page when an AI Overview was present; its December 2025 re-run on the same design measured the reduction at 58%. The aggregate landed in SparkToro and Similarweb's clickstream data: 68.0% of US Google searches ended without a click to the open web in January–April 2026, up from 60.4% in 2024.

The strategic reading: rankings and traffic have decoupled. A page can hold position one, feed the answer that satisfies the searcher, and watch its clicks fall by half. The knock-on effects across the whole organic channel — click curves, Core Web Vitals economics, what to measure instead — are collected in our SEO statistics roundup.

Which sites and formats get cited?

AI answers concentrate their trust in a remarkably short list of domains. Semrush's study of 150,000 LLM citations across ChatGPT, Perplexity, Gemini and Google AI Overviews found Reddit appearing in 40.1% of cited sources, Wikipedia in 26.3% and YouTube in 23.5% — and no other single domain cracked 5%. Community discussion, reference material and video carry the citation graph; the long tail of brand websites competes for what remains.

Format is the second filter. Wix Studio's AI Search Lab found listicles drawing 21.9% of AI citations, articles 16.7% and product pages 13.7% — three formats accounting for 52% of everything cited — with a strong neutrality bias underneath: in professional services, third-party listicles captured 80.9% of citations against 19.1% for self-promotional lists. Models reward content that compares options honestly and is structured for extraction, which is the working definition of generative engine optimization. Our free AI Visibility Checker grades any URL against exactly these citability patterns.

The volatility deserves its own line item. Semrush's multi-platform follow-up — 230,000 prompts and over 100 million citations tracked across 13 weeks — watched ChatGPT's Reddit citation rate swing from roughly 60% of responses to roughly 10% after a September 2025 retuning, and Profound recorded LinkedIn climbing from around #11 to #5 among ChatGPT's most-cited domains in three months. Model updates now reshuffle visibility the way algorithm updates once did, with no announcement and no changelog.

How much is AI referral traffic worth?

Small share, outsized intent. AI referrals average roughly 1% of total site visits across the industries Conductor studied in early 2026 — and they are growing at rates search never produced. Adobe Analytics, drawing on more than a trillion visits to 130-plus top North American retailers, measured AI-referred traffic to US retail up 393% year over year in Q1 2026.

Quality is the real story. Adobe's early-2025 data showed AI-referred visitors converting below site average; by the 2025 holiday season they converted 31% better than non-AI traffic, and by March 2026 the advantage had widened to 42%. Semrush's cohort study puts the average AI-search visitor at 4.4 times the value of a traditional organic visitor, and projects that for the digital-marketing topics it tracks, AI search could drive more visits than traditional organic by early 2028. The mechanism is straightforward: the assistant absorbs the research phase, so the person who clicks through arrives compared, qualified and close to a decision.

Even at a 1% share, a channel converting at multiples of organic deserves its own measurement and attribution. Most analytics setups still bury assistant referrals inside generic referral noise, which quietly understates the channel every quarter.

What is happening between AI crawlers and websites?

The web's old bargain — let crawlers in, get visitors back — is being renegotiated in public, and Cloudflare's network data quantifies the imbalance:

Purpose data sharpens the picture: in Cloudflare's May 2026 measurements, under 10% of AI crawler requests were search-purpose — fetching content to answer a live query with a link back — while roughly 90% served training or mixed purposes. The infrastructure layer is responding with defaults: Cloudflare began blocking AI crawlers for new domains in July 2025, launched pay-per-crawl so publishers can charge for access, and will block mixed-use crawlers on ad-carrying pages by default from September 15, 2026.

The llms.txt proposal is spreading faster than it is working. Adoption in the top 10,000 domains rose from about 1% in July 2025 to 5.6% by June 2026, and reached 8.7% of the top 1,000 sites — yet Ahrefs' server-log analysis across 137,000 domains found 97% of llms.txt files received zero requests in May 2026. Publish one cheaply; put the enforceable controls in robots.txt, WAF rules and the network edge.

What should marketers do with these numbers?

Start from Gartner's projection: brands' organic search traffic falls 50% or more by 2028 as consumers embrace generative search. The response the research supports has three parts. First, instrument AI visibility — share of answer and citation counts by platform — with the rigor once reserved for rank tracking; our free AI Brand Monitor builds the category prompt set and tracks how often each model mentions you. Second, build content shaped like what gets cited: answer-first pages, honest comparisons, original data that independent publications already quote. Third, earn presence on the domains the citation graph runs through — Reddit, YouTube, LinkedIn, third-party reviews — because roughly 90% of citations flow through territory you influence through participation and PR rather than own outright.

This is the operating work of an AI search optimization practice: crawler policy, citability engineering and share-of-answer measurement on a quarterly cadence, because the September 2025 retuning showed how fast the board resets. For the numbers on neighboring channels, our marketing statistics library collects every sourced figure in this series — the paid media statistics for auction context and the email marketing statistics for the owned-channel counterweight.