Google Search Feels Different Now. Here’s What AI Changed

The Search You Remember vs. the Search You Get

Open Google right now and search for something factual. Not your name or a website URL — something you genuinely want to know. “How long does it take to get a passport renewed.” “Why do cats purr.” “Average rent in Austin Texas.”

What you get back in 2026 looks nothing like what you got three years ago. At the top of the page, before any links, sits a block of text written by Google’s AI. It summarizes the answer. It cites a few sources in small text underneath. And for a disturbingly large number of queries, it gives you exactly enough information that you never scroll down to the actual websites.

Google calls this feature AI Overviews. It launched broadly in the United States in mid-2024, expanded through 2025, and by early 2026, it appears on roughly 48% of all tracked search queries, according to Semrush’s ongoing study. In certain categories like healthcare and education, the number is closer to 80–88%.

This is not a tweak to the interface. It is a fundamental change in the contract between Google and the internet. For twenty-five years, the deal was simple: websites create content, Google indexes it, users click through, websites earn attention and revenue. That flywheel powered the modern web. AI Overviews break it by extracting the answer and keeping the user on Google’s page.

The question is not whether this is happening. The data is overwhelming. The question is whether this is better or worse for the people doing the searching.

What the Data Actually Shows

Let me separate the measurable facts from the speculation, because there is plenty of both.

Click-through rates collapsed. Seer Interactive’s September 2025 analysis found that organic click-through rates dropped 61% for queries where an AI Overview appeared — from 1.76% down to 0.61%. Paid ad clicks fell even harder, crashing 68%. When the AI gives you the answer, there is less reason to click on anything, including the ads Google sells.

Zero-click searches are now the majority. As of late 2025, roughly 58–60% of all Google searches end without a single click to any external website. For queries that trigger an AI Overview specifically, the zero-click rate climbs to 83%. Users see the AI summary, get what they need, and close the tab.

But the picture is more complicated than “AI killed clicks.” A Semrush study found that when researchers compared the same search terms before and after AI Overviews appeared, users who saw the overview actually clicked slightly more in some categories. The hypothesis is that the overview acts as a primer: it gives users enough context to formulate a better follow-up query or to know which result is worth clicking. Whether this holds broadly or only for certain query types remains an open question.

Some publishers benefit enormously. Brands and websites that get cited inside an AI Overview earn 35% more organic clicks than they did before, because the citation acts as a trust signal. The problem is that only a handful of sources get cited per overview, so the benefits concentrate at the top while everyone else loses traffic.

The User Experience, Honestly Evaluated

Here is where I will be blunt about something the SEO industry does not want to hear: for many queries, AI Overviews are genuinely better for the user.

If I search “how many ounces in a cup,” I do not want to visit a food blog, scroll past a life story about the author’s grandmother, dodge three pop-up ads, and finally find the number buried in the fourth paragraph. I want the number. The AI gives me the number. This is an improvement.

For simple factual queries, definitions, unit conversions, and straightforward how-to questions, the AI summary eliminates friction that nobody enjoyed. The old experience of clicking into a content-farm article optimized for Google rather than for readers was already broken. AI Overviews did not break search for these queries. They fixed it.

But the story reverses for anything requiring nuance, multiple perspectives, or personal judgment. Search “best laptop for video editing” and the AI Overview gives you a synthesized answer that sounds authoritative but flattens the complexity. It does not tell you that the recommendation changes depending on whether you edit in Premiere or DaVinci Resolve, whether you need portability, or whether you are budget-constrained. It picks a few sources, blends their opinions, and presents a confident paragraph that hides the trade-offs.

For medical queries — which trigger AI Overviews 88% of the time in healthcare searches — the stakes are higher. An AI summary of symptoms and treatments might be directionally correct but miss the edge case that matters to you specifically. The old model, where you clicked through to WebMD or Mayo Clinic and read their full article, at least gave you the context to evaluate the information. The AI summary strips that context away.

The honest assessment: AI Overviews made simple searches better and complex searches worse. The problem is that Google applies the same feature to both.

What This Means for the People Who Make the Internet

If you run a website that depends on organic search traffic, the landscape has shifted permanently. This is not a temporary algorithm fluctuation. It is a structural change in how Google distributes attention.

The hardest-hit category is what the industry calls informational content. Articles that answer factual questions, explain concepts, or provide step-by-step instructions — the bread and butter of content marketing for fifteen years — are being summarized directly in the search results. The AI Overview effectively replaces the click. A site that once received 100,000 monthly visitors from “how to” queries might see that number cut in half, not because their content got worse, but because Google no longer needs to send users there.

The publishers who are surviving, and in some cases thriving, share a few common traits. They produce content the AI cannot easily summarize: original research, proprietary data, interactive tools, and opinions backed by demonstrated expertise. They build direct relationships with their audience through email lists and communities rather than relying solely on Google as a distribution channel. And they optimize for being cited inside the AI Overview rather than just ranking in organic results.

Small publishers and independent bloggers face the worst of it. They lack the brand recognition that makes Google’s AI cite them as a source, and they depend disproportionately on long-tail informational queries that AI Overviews absorb most completely. A travel blogger writing “best time to visit Kyoto” is competing against an AI that synthesizes dozens of articles and presents the answer without attribution to any single source.

The economic logic is uncomfortable but straightforward. Google’s revenue depends on keeping users on its platform. AI Overviews increase time on Google’s page. The content creators whose work the AI summarizes receive less traffic. The value flows upstream to Google, and the people who created the underlying information get a smaller share of the attention.

Where This Goes From Here

Google is not going to reverse course. AI Overviews are expanding, not contracting. The company reported that users who see AI Overviews engage more with search overall, and from Google’s perspective, any feature that increases engagement justifies its existence. Expect AI Overviews to appear on an increasing share of queries through 2026 and beyond.

The more interesting development is AI Mode, which Google began testing in early 2026. This is a conversational interface layered on top of search — more like ChatGPT than traditional Google. Instead of a single AI summary, users can ask follow-up questions and have a multi-turn conversation with Google’s AI, with web results woven into the responses. If AI Overviews reduced clicks by 61%, AI Mode threatens to make the traditional results page irrelevant for entire categories of queries.

Competition is forcing the issue too. ChatGPT, Perplexity, and Microsoft’s Copilot all offer AI-powered search that bypasses Google entirely. For the first time in two decades, Google faces genuine competition in its core business. The response has been to accelerate AI integration, not slow it down. If Google does not provide AI answers, users will get them elsewhere.

For users, the near-term trajectory is clear: search will feel faster and more convenient for routine queries. You will get answers without clicking. You will ask follow-up questions without starting a new search. The friction of navigating the web will decrease.

The cost of that convenience is less visible but real. Fewer clicks mean less revenue for content creators. Less revenue means less incentive to create the original content that AI systems need to generate useful answers. If the current trajectory continues without meaningful changes to how content creators are compensated, the quality of the information feeding these AI systems will eventually degrade. The AI will summarize thinner and thinner source material, and the answers will get worse without anyone noticing the slow decline.

That is the trade-off nobody at Google talks about publicly. The search experience got better today by borrowing against the content ecosystem that makes it work tomorrow. Whether that debt comes due, and how soon, is the question that will define the next chapter of the internet.

Frequently Asked Questions