Google just made a big claim about AI Search, and the reaction from the SEO and publishing world tells a very different story than the one Google is telling.
A senior Google executive overseeing Knowledge and Information recently shared an update on how people are actually using AI inside Search. The message was simple on the surface: AI hasn’t killed search, it’s made people search more, and AI features inside Search are sending real traffic back to websites, not keeping users trapped inside Google’s own answer boxes.
According to the post, Google continues to send billions of clicks to the web every day through traditional Search, and now claims to be sending billions of additional clicks to websites every week through AI features in Search alone. The framing was optimistic: AI in Search connects people to websites rather than replacing them, and the company says it’s still improving that experience.
The Core Claim, Broken Down
The statement rests on three ideas stacked together.
First, that early fears about AI making search obsolete haven’t played out. People assumed that if AI could just answer a question directly, nobody would bother searching anymore. Google says the opposite happened. Lowering the friction to ask a question apparently led to more questions being asked, not fewer.
Second, that Search as a whole still sends billions of clicks to websites every day, a number Google has cited before.
Third, and this is the new part, that AI features specifically, separate from traditional blue-link search, are now sending billions of clicks to websites every week.
Read quickly, this sounds like reassurance for anyone worried that AI Overviews and AI Mode are quietly strangling referral traffic. Read carefully, and it raises more questions than it answers, which is exactly what the SEO community pointed out almost immediately.
Why Nobody in SEO Is Fully Buying It
The response underneath the post was long, technical, and mostly skeptical. A few themes kept coming up again and again, from marketers, consultants, product managers, and people who work directly with search data every day.
Aggregate Numbers Hide Individual Losses
The most repeated point was simple: big aggregate numbers can go up while individual websites still lose traffic. Google operates at a scale where “billions of clicks” is almost a meaningless unit on its own. What actually matters to a business owner or publisher is whether their specific site is getting more or fewer visitors than it was getting a year ago, and for a lot of publishers, especially in content-heavy niches, the honest answer has been fewer.
Several commenters made the same underlying argument in different words: an aggregate figure can rise across the entire internet while specific newsrooms, blogs, and service businesses still watch their own visibility and revenue shrink. Scale can hide decline. It doesn’t disprove it.
Nobody Can See This Data Themselves
The second major theme was about transparency. If Google is genuinely sending billions of clicks to websites every week through AI features, the natural question is why that traffic isn’t clearly broken out anywhere inside Google Search Console. Website owners can see impressions and clicks from regular search results, but AI Overviews and AI Mode traffic isn’t cleanly separated in a way most site owners can verify on their own end.
More than one person pointed out that competing platforms already provide more granular referral data than Google does here, which made the claim feel more like a talking point than something the industry can actually audit. When a company is the only one holding the full dataset and also the one making the claim, healthy skepticism is the reasonable default, not the exception.
The Fan-Out Problem
A more technical objection came from people who understand how AI Mode actually processes a query. When someone asks a single question inside AI Mode, Google doesn’t necessarily run one search behind the scenes. It can run several related searches to build out a fuller answer, sometimes referred to as query fan-out. That raises a genuine measurement question: if one user question triggers ten background searches, does that count as one search or ten in Google’s reporting? The same question applies to clicks. Nobody outside Google currently has a clean answer, which makes the headline number harder to interpret at face value.
Is AI Additive, or Just Counted Twice?
Another recurring question was whether AI-driven clicks are a genuinely new source of traffic, or whether they’re simply traditional search clicks being recategorized and counted a second time under a different label. If someone bounces between a regular search result and an AI-generated summary before finally clicking through, which bucket does that click land in, and is it being counted once or in both categories? Without a clear methodology being shared publicly, this distinction stays unresolved, and it directly affects whether the “billions of clicks” framing represents real incremental value or just relabeled activity.
What Percentage of Searches Actually Convert to a Click?
One of the more pointed questions raised in the comments asked for the denominator, not just the numerator. If there are trillions of searches happening per year, what fraction of those searches actually result in a click through to an external website, and how has that percentage changed since AI features launched? A raw click count without that context can rise even while the click-through rate per search quietly falls, which would still be bad news for anyone relying on search traffic to run a business.
Publishers Are Reporting the Opposite Experience
Away from the abstract data debate, some of the most direct pushback came from people who create content for a living. Multiple commenters described AI-generated answers misrepresenting brands, recipes, photos, and videos at a scale they hadn’t experienced before, essentially using their work to build an answer while reducing the incentive for a user to ever visit the original source. From that seat, the billions-of-clicks framing doesn’t match daily reality, where referral traffic to original content creators has been trending down, not up.
A Broader Behavioral Shift Worth Noting
Not every comment was purely critical. A few people raised a genuinely interesting observation that cuts in Google’s favor: AI hasn’t necessarily reduced curiosity, it’s lowered the barrier to asking a question in the first place. That means people may now be exploring topics they never would have typed into a traditional search bar before, simply because asking feels lower effort. If true, that supports Google’s claim that search volume itself is growing. The unresolved part is whether that expanded curiosity is translating into meaningful traffic for the websites actually answering those questions, or whether the AI layer is absorbing most of that value before a click ever happens.
What This Actually Means If You Run a Website
Strip away the debate and a few practical realities remain, regardless of which side of the argument turns out to be right.
Aggregate platform claims don’t tell you anything about your own site.
The only data that matters for decision-making is what’s happening inside your own Search Console property, your own analytics, and your own lead volume. If a platform-wide announcement doesn’t match what your own numbers show, trust your own numbers.
AI visibility is becoming its own discipline.
Whether or not Google is fully transparent about the click data, it’s increasingly clear that showing up correctly inside AI Overviews, AI Mode, and other AI-generated answers is a separate skill from traditional ranking, sometimes called AEO or GEO. Structuring content so it can be accurately cited, rather than paraphrased or misrepresented, is becoming as important as traditional on-page SEO work.
Diversifying traffic sources is no longer optional.
With multiple commenters noting real referral shifts toward alternative search engines and AI-native tools, relying on a single traffic source for a business is a riskier bet than it used to be. Building direct relationships with an audience through email, community, or direct navigation is a hedge against exactly this kind of ambiguity.
Demand real data before changing strategy based on a headline.
Announcements like this tend to shape industry conversation for weeks. Before reallocating budget or shifting a content strategy in response to a claim like “billions of clicks weekly,” it’s worth waiting to see whether any verifiable, third-party data eventually supports it. So far, that verification hasn’t arrived, and the loudest voices closest to the data are the ones asking the hardest questions.
The Bottom Line
Google’s claim isn’t necessarily false. It’s unverifiable, and that’s the actual problem the industry has with it. A platform holding all the data, making a positive claim about itself, without giving outside site owners a way to confirm it, is always going to be met with the kind of skepticism this post received. Until Search Console or an equivalent tool actually breaks out AI-driven referral traffic in a way publishers and SEOs can see for themselves, the honest position for anyone running a website is to keep watching their own numbers closely, keep building visibility across more than one channel, and treat big aggregate announcements as a talking point, not a strategy.
