Google Created Billions of New AI Searches. Semrush Proved Nobody Owns Them.

Two things happened this month that most founders missed entirely. Google's Q2 earnings revealed that AI is creating billions of searches that never existed before. And a Semrush study of 50,000 brands proved that 85% of the categories those searches land in have no brand leader. That gap is the single biggest opportunity in digital marketing right now, and the old playbook does not chart it.

85% of AI Categories Are Unclaimed. The High-Demand Ones Are Even Worse.

Semrush analyzed 1,094 U.S. categories across ChatGPT using five standard buyer prompts: definitions, comparisons, alternatives, use cases, and purchase decisions. A brand qualified as a category owner by appearing in at least four of five responses while leading the next competitor by 5 percentage points or more.

The results should make you rethink everything you assume about AI visibility.

Only 15.2% of categories had a clear brand owner. Another 31.2% showed an emerging leader. And 53.7% were completely unsettled, with no brand appearing consistently in AI responses at all.

Now here is the part nobody is talking about. The highest-demand categories, the ones that account for 98% of total AI search volume, had even lower ownership. Just 11.3%.

Almost all of the search activity in AI is flowing through categories nobody owns.

Your SEO Metrics Are Not Predicting What You Think They Are

I have spent nearly a decade watching brands pour money into domain authority and organic traffic rankings. The Semrush data shows those metrics predict AI brand ownership about as well as guessing.

Authority Score: owners led in 52.5% of category comparisons. Organic traffic: 48.4%. Both are statistically indistinguishable from a coin flip.

Branded search volume was the only metric that showed meaningful correlation at 55.7%. Still not overwhelming, but it was the only one that registered at all.

The most damning number: only 21% of the most-cited domains in ChatGPT were also the most-mentioned brands. The domains getting linked and the brands getting named are two different lists. If you are measuring citations and assuming you are measuring brand visibility, you are measuring the wrong thing.

The entire measurement apparatus most companies rely on tells them almost nothing about whether they win in AI. 45% of marketers cannot even measure AI brand visibility. Only 9% can measure all the relevant metrics: per-platform mentions, citations, mention-source overlap, and cross-platform consistency.

36 Brands Win Everywhere. Everyone Else Is Platform-Dependent.

Semrush's AI Visibility Index tracked 1,200+ brands across ChatGPT, Gemini, Google AI Mode, and AI Overviews from January through April 2026, analyzing 126 million U.S. prompts.

Thirty-six brands maintained top-100 visibility across all four platforms, every month. YouTube. Amazon. Facebook. Apple. Walmart. Netflix. The companies that were already category-defining before AI existed.

Everyone else is platform-dependent. A brand might show up consistently on ChatGPT and not exist on Gemini. Only three brands, YouTube, Amazon, and Facebook, appeared in the top 10 on every single platform.

Here is why that matters for you: ChatGPT averages 3.3 brand mentions per response with 15.4 sources. Google AI Mode averages 5.2. Every response is a winner-take-most event, and most of those events are happening in categories where the winner has not been decided yet.

Google Just Told You the Pool Is Growing

During Google's Q2 2026 earnings call, SVP Philipp Schindler said AI Max is matching ads to searches that were "too complex or ambiguous for traditional keyword targeting." Over 500,000 advertisers have adopted it. Conversions lifted 15%.

Read past the ad numbers. Google is admitting that AI search creates entirely new types of queries. Not searches people would have typed differently. Searches that would not have happened at all.

Queries like "what's the best CRM for a 12-person agency that also does client reporting" do not fit in a keyword box. They fit in a conversation. And AI handles them. These new discovery moments are expanding the total pool of searchable intent, and the default owner of those new moments is nobody.

Combine that with the Semrush data: the territory is growing AND it is empty.

First Movers Keep Their Position. But the Margins Are Thin.

One more number from the Semrush study that should sharpen the urgency.

Once a brand became a clear category owner, it retained first place in 90.4% of month-over-month comparisons. That is real durability. First-mover advantage in AI categories is not theoretical.

But the fragility sits in the margin. Topics where leaders held their position had a median lead of 2.9 percentage points. Topics where leadership changed had a median lead of just 1.3 points. Category leaders changed nearly 2,000 times during the study period, almost always where leads were narrow.

So the advantage compounds, but only if you build it with enough separation. A thin lead gets flipped. A decisive lead sticks.

What I Would Do With This Data

The situation is binary and it is time-sensitive.

Right now, AI discovery is the digital equivalent of unclaimed territory. 85% of it has no owner. The search volume is growing because AI is creating query types that were previously impossible. The metrics most brands use to measure "winning" correlate less than a coin flip with the thing that actually matters.

Here is the move.

Stop treating AI visibility as an extension of SEO. The 48.4% organic traffic correlation proves it is not. These are different systems running on different logic. The teams that integrate SEO and AI search workflows report 81% more traffic and leads than those running them separately.

Measure what actually predicts ownership. Branded search volume is the only metric with real correlation to category leadership. Build the kind of brand recognition that makes buyers search for you by name, because that is what AI engines recognize as authority.

Claim a category before someone else does. The Semrush data covers January through June 2026. By the time this study runs again, a significant number of those 53.7% unsettled categories will have owners. The question is whether your brand is one of them.

This is what Machine Relations exists to solve. Not traditional PR. Not traditional SEO. The discipline of building the source material that AI engines extract, cite, and recommend when a buyer asks a category question with your name nowhere in the prompt.

The territory is real. It is growing. And right now, it is yours for the taking.

Or it is not.

FAQ

How does AI brand ownership differ from traditional search rankings?

In traditional search, you rank for keywords. In AI, you either get mentioned by name when a buyer asks a category question, or you do not exist. Semrush found that only 21% of the most-cited domains are also the most-mentioned brands. Citation and brand mention operate on completely different logic, which is why strong traditional SEO does not guarantee AI visibility.

Can smaller brands compete for AI category ownership?

Yes. The data actually favors them in one critical way: only 11.3% of high-demand categories have clear owners. The vast majority of AI search volume flows through categories where no brand dominates, creating a first-mover window for companies that build topic-level authority before larger incumbents wake up to the opportunity.

What metrics should founders track for AI visibility instead of Authority Score?

Branded search volume is the only metric with statistically significant correlation to AI category ownership at 55.7%. Authority Score (52.5%) and organic traffic (48.4%) are statistically no better than chance. Track how often your brand gets mentioned by name in AI responses across ChatGPT, Gemini, AI Mode, and AI Overviews, not how many links point to your domain.