Google AI Overviews Statistics September 2026: Reach, CTR Impact & Regulation

How many people see Google AI Overviews and AI Mode, how much they cut publisher click-through rates, and how the UK-forced opt-out has now rolled out worldwide.

Paul Boudet

Paul Boudet · Co-founder

Key takeaways

  • AI Overviews now reach 2.5 billion+ monthly users — a figure Google itself confirmed for the first time in an August 31, 2026 update to its own blog, up from the 2 billion it had cited since May 2025. AI Mode, the more conversational Search experience, crossed 1 billion monthly users by May 2026 and Google reaffirmed that same milestone in the August 31 update — up from 75 million daily active users less than a year earlier.

  • Query coverage estimates vary wildly by tracker — from roughly 20% to 50% of searches — but nearly every study agrees coverage has expanded well beyond informational queries into commercial, transactional, and even shopping searches over the past year.

  • Multiple independent studies confirm AI Overviews cut publisher click-through rates — Pew Research, Ahrefs, and Seer Interactive all found meaningful declines, and a randomized field experiment provided the first causal (not just correlational) evidence, finding a 39.8% drop in outbound clicks. A newer wrinkle: branded queries that trigger an AI Overview actually see an 18% CTR increase, suggesting AIOs redistribute clicks toward cited brands rather than suppressing them uniformly.

  • The publisher opt-out the UK forced Google to build is no longer UK-only. After the CMA's June 3, 2026 "world first" ruling and an initial UK-only test, Google rolled the same Search Console opt-out control out to all websites worldwide as of August 31, 2026 — well ahead of the CMA's own December 2026 deadline for the feature. On September 1, 2026, Google confirmed it's offering that same global opt-out as a move to resolve a separate EU antitrust probe into how it uses publisher content for AI Overviews.

  • Google's Search revenue keeps growing despite the click declines — up 19% year over year to $60.4 billion in Q1 2026, then up 17% year over year to $63.3 billion in Q2 2026 (reported July 22). Alphabet's Q3 2026 results aren't due until late October or early November, so Q2 remains the latest disclosed figure. Alphabet maintains AI features are "driving Search growth," a claim publishers and several independent studies directly dispute.

  • Google announced five changes to AI Overviews in May 2026 — including subscription labels and more prominent citations — widely read as a response to mounting publisher pressure, even though Google's own announcement never mentions clicks or lost traffic.

Google AI Overviews are a fundamentally different animal from ChatGPT or Perplexity: nobody chooses to open an "AI Overviews app." It's a feature embedded directly in Google Search results, seen by billions of people who never consciously decided to use an AI product. That makes its statistics less about user growth and more about two questions that matter enormously to anyone who publishes content or runs ads: how many searches does it touch, and how much of the click-through traffic that used to flow to websites is it keeping for itself? For brands, understanding both is central to ranking in AI Overviews and to tracking AI Overview visibility as the feature keeps expanding.

Here is a full breakdown of the numbers as of September 2026.

How many people see AI Overviews and AI Mode?

Google now says AI Overviews reach 2.5 billion+ monthly users, available in more than 200 countries and territories and 40+ languages since the feature's May 2025 international expansion. That's an update from the 2 billion figure Google had cited since first disclosing it in mid-2025 — third-party trackers had already been citing estimates around 2.5 billion for months, but Google's own blog only confirmed the higher number in an August 31, 2026 update. Because AI Overviews are embedded in ordinary Search results rather than a separate destination, this "reach" figure counts anyone who ran a search that happened to surface a summary — not necessarily someone who deliberately sought out an AI answer.

AI Mode, Google's more conversational, multi-turn Search experience, is a separate and smaller product that has scaled quickly:

Milestone

Figure

Daily active users, ~2025

75 million

Monthly active users, US + India

100 million+

Monthly active users, worldwide (Google I/O, May 2026)

1 billion+

Monthly active users, worldwide (Q2 2026 earnings call, July 22)

1 billion+ (reaffirmed)

Monthly active users, worldwide (Google blog update, August 31, 2026)

1 billion+ (reaffirmed again)

At Google I/O 2026, Google expanded "Personal Intelligence" in AI Mode — which can draw on a signed-in user's Gmail and other Google app context — to nearly 200 countries and territories across 98 languages, with no subscription required. Google says AI Overviews drive a 10%+ increase in Search usage for the query types that display them, meaning people come back to do more of that type of search once they've seen an AI answer for it. On the Q2 2026 earnings call, Sundar Pichai reiterated the 1 billion+ AI Mode figure and said Search usage hit a record high during the World Cup, with Google framing itself as sending "billions of clicks to websites every week" through AI features in Search — a framing publishers and independent CTR researchers continue to dispute (see below).

Which searches trigger AI Overviews?

Query-coverage estimates differ enormously depending on who's tracking and what keyword set they use — a real limitation of this data, not just noise:

  • ~20% of US searches, per Seoprofy's tracking (19–20% range across the US, Canada, and UK)

  • ~30% of US searches, per DemandSage (27.75–28.66% depending on state)

  • ~48% of tracked queries, per BrightEdge — up roughly 58% year-over-year from February 2025

  • ~50% of US queries, per DemandSage's headline figure

What's consistent across every tracker is the direction and shape of the expansion: AI Overviews started out almost entirely on informational queries and have since spread into commercial and transactional territory. Semrush's analysis of 10 million+ keywords found informational queries fell from 91.3% of AI Overview triggers in January 2025 to 57.1% by October 2025, while commercial queries rose from 8.15% to 18.57% and transactional queries from 1.98% to 13.94%.

Trigger rates also vary sharply by query shape:

Query type

Trigger rate

Comparison ("X vs Y")

95.4%

Reviews

86.3%

Question-format

85.9%

"Near me"

76.9%

Health / relationship topics

~60%

Shopping queries

14.0%

Local intent

~7%

Single-word queries

27.3%

Source: Seer Interactive, DemandSage, SearchEngineLand. Longer, more specific queries trigger far more often — seven-plus-word queries at roughly 46%, versus under 10% for one-word queries — and about 81% of AI Overview impressions happen on mobile.

When Google does cite sources, user-generated platforms dominate: YouTube, Reddit, and Wikipedia are consistently the top three citation domains across independent studies, ahead of traditional publisher sites.

The publisher impact: what the click-through data shows

This is the most contested — and most studied — question in AI search. Multiple independent research efforts, using different methods and samples, have all found the same basic pattern: showing an AI Overview reduces the odds a user clicks through to a website.

  • Pew Research Center (July 2025) tracked 68,000 real search sessions and found users clicked a traditional result 8% of the time when an AI Overview appeared, versus 15% without one — a 47% relative decline. Only 1% of users clicked a link inside the AI Overview itself.

  • Ahrefs found the top-ranking organic result lost 34.5% of its expected clicks when an AI Overview appeared (April 2025, 300,000 keywords). By May 2026, that had worsened to a 58% average CTR decline for top-ranking pages — with position-two results losing about half their clicks and even position-ten results down nearly 20%.

  • Seer Interactive's longest-running tracker (42 organizations, 25M+ impressions, June 2024–September 2025) found organic CTR on AI Overview queries fell 61% (1.76% → 0.61%) and paid CTR fell 68% (19.7% → 6.34%). A 2026 update on a different sample showed some recovery — organic CTR climbing from a floor of 1.3% in December 2025 back to 2.4% by February 2026 — suggesting the decline may be stabilizing rather than continuing indefinitely.

  • A randomized field experiment (1,065 desktop Chrome users, published on SSRN in April 2026 and revised on June 17) provided the first causal rather than correlational evidence: showing an AI Overview caused outbound organic clicks to fall 39.8% and zero-click searches to rise 34.5%, while clicks on paid results stayed essentially flat.

  • DMG Media (MailOnline, Metro), in a submission to the UK's competition regulator, reported worst-case declines as steep as 89% for certain searches — desktop CTR falling from 25.23% to 2.79% when an AI Overview appeared above a visible link, with mobile down 87%.

  • Chartbeat data cited by Press Gazette found publishers worldwide have seen roughly 33% fewer search engine referrals overall since AI Overviews rolled out broadly (38% in the US specifically).

  • Not every query type is hurt equally. Digital Applied's March 2026 analysis found branded queries that trigger an AI Overview see an 18% CTR increase rather than a decline — evidence that AI Overviews can redistribute clicks toward the specific brands they cite, even as they suppress clicks overall for non-branded, informational queries.

Google disputes the framing, not usually the raw numbers. CEO Sundar Pichai and Search VP Liz Reid have both argued publicly that AI Overviews mainly eliminate "low-value" clicks — searches where a user wanted a quick fact, not a page visit — and that overall Search usage and query volume are at all-time highs as a result. Independent studies haven't settled that dispute either way, but the consistent direction of the click-through data across five separate research efforts, using five different methodologies, is difficult to dismiss as noise.

None of this makes AI Overviews avoidable — it makes being the cited source inside one more valuable, not less. See our practical guide on how to optimize for AI Overviews for the specific tactics that improve citation odds.

Regulators are starting to intervene

Publisher pressure has moved from complaints to enforcement. Independent Publishers Alliance and the campaign group Foxglove filed a complaint with the UK's Competition and Markets Authority (CMA) in July 2025, later joined by more than 20 major publishers.

In one of that month's biggest AI search stories, on June 3, 2026, the CMA — acting under the UK's new digital markets competition regime, and citing Google's 90%+ share of UK search — announced what it called a "world first": Google must let publishers opt out of having their content used to power AI Overviews and AI Mode specifically, without losing their normal search rankings or visibility. Google must also let publishers opt out of having their content used for AI model "fine-tuning," and must properly attribute publisher content in AI-generated results with clear links. Google had already begun testing opt-out tooling ahead of the ruling, and said it would keep working with regulators as "user preferences evolve."

The CMA's order set a phased implementation timeline — main publisher opt-out controls by December 2026, with more granular page-level "grounding" controls following by March 2027 — but Google moved faster than that deadline required. It began testing the new Search Console toggle with a subset of UK website owners shortly after the June ruling, then, per a footnote added to its own announcement, rolled the control out to all websites worldwide on August 31, 2026 — more than three months ahead of the CMA's December deadline. Once opted out, a site stops receiving any traffic or impressions from AI Overviews, AI Mode, or AI Overviews in Discover, and the toggle is explicitly excluded from affecting ranking in ordinary search results. Google also began rolling out new Search Console insights alongside the opt-out: impression metrics and country-level data on which pages appear in AI responses, with more metrics promised over time.

The order matters beyond the UK because it's the first time any regulator has forced a structural separation between "appearing in Google Search" and "being used to power Google's AI features" — a distinction publishers had been asking for since AI Overviews launched, since ordinary robots.txt rules can't block just the AI summary without blocking search visibility entirely. Italian news publishers have separately demanded a similar investigation, and the European Union opened its own antitrust probe into Google's AI Overviews and AI Mode in December 2025, examining whether Google is "abusing its dominant position as a search engine to impose unfair trading conditions on publishers." On September 1, 2026, reporting indicated Google is now offering that same global opt-out specifically as a proposed remedy to resolve the EU probe — extending what started as a UK-only fix into Google's answer to a second, separate regulatory threat.

Separately, on May 6, 2026, Google announced five product changes to AI Overviews and AI Mode: labeling when a summary draws on a publication a user subscribes to, more prominent inline citations "next to the relevant text," and a new "Further Exploration" links section. Google's own announcement frames these purely as usability improvements and doesn't mention clicks, traffic, or publisher pressure — but independent commentary widely reads the timing, arriving just weeks before the causal CTR study's revision and the CMA ruling both landed, as a response to exactly that pressure.

Revenue and business impact

Whatever AI Overviews are doing to individual publishers, they haven't dented Google's own Search revenue. In Q1 2026, Alphabet reported:

  • Google Search & Other revenue: $60.4 billion, up 19% year over year

  • Total Alphabet revenue: $109.9 billion, up 22% year over year

  • Google Cloud revenue: over $20 billion for the first time in a quarter, up 63% year over year, with backlog nearly doubling to over $460 billion

  • Paid subscriptions across Google products: 350 million, led by YouTube and Google One

Growth continued into Q2 2026 (reported July 22), with revenue accelerating rather than slowing. Alphabet's Q3 2026 results are expected in late October or early November, so these Q2 figures remain the most recent as of September:

  • Google Search & Other revenue: $63.3 billion, up 17% year over year

  • Total Alphabet revenue: $119.8 billion, up 24% year over year

  • Google Cloud revenue: $24.8 billion, up 82% year over year

Google's own framing was consistent quarter to quarter: Pichai told Q1 investors that "AI experiences [are] driving usage, queries [are] at an all-time high, and [there's] 19% revenue growth" in Search, and repeated the same argument in Q2 — that AI Overviews expand the overall pie rather than cannibalize the clicks publishers used to get, and that Search advertising in particular hasn't visibly suffered from the shift to AI-generated answers. Independent CTR research, summarized above, complicates that story without fully resolving it: query volume and Google's own revenue can rise even as the specific traffic publishers receive per search falls, because the two are measuring different things.

How AI Overviews compares to standalone AI products

By sheer reach, AI Overviews' 2.5 billion+ monthly users still dwarf any standalone chatbot — though the gap has narrowed since June. ChatGPT itself crossed 1 billion weekly active users in late July 2026, and the standalone Gemini app separately hit 1 billion monthly active users on August 11, 2026, up from 950 million reported a month earlier on Alphabet's Q2 earnings call. But as covered in more depth in our ChatGPT statistics post, none of these comparisons are apples-to-apples: an AI Overviews "user" ran a Google search that happened to surface a summary, while a ChatGPT or Gemini user deliberately opened an app and typed a message. Intent and engagement depth differ enormously between the categories, even though Google's numbers are still the largest by a wide margin.

Sources

Published Jun 21, 2026

Updated Sep 28, 2026

Genlook

How Genlook became #1 in ChatGPT

🪄 Case Study