Only 28% of Americans Trust Search AI – And That Gap Is Opening for Your SEO

Search engines are not losing the trust battle to AI chatbots. They won it by more than 40 points, and nowhere is that margin wider than in the United States.
In YouGov’s live stream on July 8, “The New Search Journey, How AI Is Changing Online Discovery.” Host Brian Reitz walked experts Clifton Mark and Jade Vasquez through a new survey of 19 markets on how consumers use search engines and AI assistants, where they start different information activities, and what makes them trust an AI-generated answer enough to act on it. Vasquez, who has a master’s degree in computational social science from UC San Diego and often applies that lens to sports and technology audiences, and Mark, a senior business data journalist who spent years hosting a podcast called “Good in Theory,” were on hand to explain why theory and practice differ. I signed up because of the article, but stayed because the report answered a keyword question the tools don’t know. Search volume tells you what people are typing; this survey tells you who is writing it, and why they don’t trust the answer.
I have spent 25 years arguing that market research and search data are two different tools for measuring two different things. This report is the clearest evidence I’ve seen this year of why SEO practitioners need both.
Article Topic No SEO Wants Help
Here’s a number that should rebalance a lot of 2026 planning. Among the 19 markets YouGov researched, the US had the lowest rate of AI-assisted searches of any country in the study, at 48%. Compare that to 89% in India, Indonesia, and the UAE. Even Great Britain, the next most cautious market, sits at 54%. Americans aren’t just slow to discover AI search. They are global industries.
Themba tells a similar story. Only 28% of US online searchers say they trust information from an AI assistant, compared to 70% who trust a search engine and 76% who trust a maps or navigation app. AI assistants are more than just social media, which is not something any company or brand wants its citation strategy to maintain.
Mark Fantino, senior vice president of YouGov America, put the power in the foreword of the report. “They just want to answer you,” he wrote of AI assistants, before making a very important point for anyone building a content strategy around them. The catch, as it is not in the draft, is that AI may save people steps, but people still want receipts, which means links to the source, official sites, something to be verified against him.
I think one word, receipts, sums up SEO better than most I’ve read this year on generative engine optimization (GEO).
Where the Search Really Begins, Task By Task
The report breaks down where humans are starting seven general knowledge tasks, and the pattern undermines the assumption that AI is already the default for anything. Search engines lead all researched work. By asking a specific question, the case of using AI assistants that are said to be designed to win, 69% of Internet searchers still start with a search engine and only 16% start with AI. For research products, 62% search engine versus 4% AI. Buying products, 50% compared to 2%.
One place AI shows real strength is inside the ride, not in front of it. Among people who use AI assistants to search, only 16% call AI their first real choice. Thirty-two percent use it after trying other sources first. Another 27% only use it for specific questions when they already suspect there is a definite answer. AI, in other words, acts as a second opinion, not a primary one.
That’s important because it reframes what AI visibility is really worth. If a brand’s content is cited within an AI response that the user only accesses after they have searched elsewhere, the AI citation is not the same as the search result. Ride on it.
What’s Behind AI Actually Answers
This is the part of the report that I found most helpful, and the part that Reitz spent real time unpacking with Mark and Vasquez during the live stream. When an AI assistant answers a search query, 22% of AI searchers say they tend to click on the provided links anyway. Another 16% compare the answer to other applications. Only 17% said they usually stop searching after receiving an AI response. Reduce that to frequency, daily AI searchers, and the click-through rate rises to 33% while the stop rate remains low at 17%, meaning that people who rely heavily on AI are also the most likely to treat its answer as the end of the search.
That’s the opposite of the “zero-click apocalypse” framework that has dominated much SEO commentary this year. People who use AI assistants frequently are not the most likely to accept AI feedback at a significant rate. They are the most likely people to confirm it.
This is where I will put the position rather than the parentheses. The horror of AI search in this field is aimed at the wrong person. The threat was never that AI chatbots would replace search traffic, and the real risks are smaller and more solvable. That your product is not the source that the AI cites, and it is not the source that the search engine clicks on to verify it. Dissolve citation and verification together, and the zero click frame usually stops working for you.
Signs of Trust That Really Move People, and Those That Don’t
YouGov asked both AI and non-AI searchers what would increase their trust in an AI-generated answer. Among people who already use AI to search, 16% said clear links to sources would be more helpful, 15% pointed to an answer from an official source, and 14% wanted to see more sources side by side.
Now look at non-AI search engines, which SEO teams desperately want to convert. Forty-nine percent of them said that none of the trusted brands on the list would change their minds. There is no one. That’s a surprisingly high number, and it confirms something the report’s authors specifically said. Transparency features are much better at deepening trust among people who already use AI than converting people who don’t.
This is where historical parallels should be made, carefully. In the early 2000s, ecommerce faced almost the same problem. Consumers didn’t refuse to shop online because checkout pages lacked features. They refused because no one had confirmed that the transaction was safe. What filled that gap was not a clever copy. It was third-party authentication, padlock icons, escrow-like guarantees, predefined return policies, the digital equivalent of a receipt. AI search is at the same stage as e-commerce was around 2002. Correction is not the best prose. It is physical evidence.
Personalization runs through the same wall. Sixty-eight percent of non-AI searchers say they are comfortable with AI assistants using their data to compile answers, and even among people who already use AI, only 31% are comfortable with it, and only if they can control it or turn it off. If your GEO strategy hinges on the assumption that personalization will be the limit that attracts skeptics, this data says otherwise.
Who’s Really Driving Growth, and Who’s Not
Productivity data reinforces all of this rather than contradicting it. Fifty-four percent of Americans look up information online every single day, and a third of Gen Z and Millennials do it six or more times every day. Young adults also carry a wider platform as legitimate search tools. Millennials lead the use of AI assistants in search at 33%, ahead of Gen X at 22% and Baby Boomers+ at 20%.
But the growth story for the next 12 months is not about converting new users. It’s about deepening usage among people who are already there. Fifty-three percent of the average AI searchers expect to use AI even more in the next year. Among people who do not use AI for search at all, only 4% expect to start, and 72% expect to switch completely. The authors of this report call this “deep engagement rather than extensive non-user conversion,” and I would put it bluntly. The AI search market in the US is not expanding outward. It’s socializing internally, among a small group of people who would always be your first adopters anyway.
What You Really Should Do About This, Starting This Week
None of this is an argument for ignoring AI search. It’s an argument to stop budgeting for it as a replacement for your search strategy, and start building it as a layer on top of another that we still have to work on ourselves.
First, keep investing in classic search basics like the primary channel, not the historical one. Eighty-six percent of online searchers have used a traditional search engine in the past 30 days, and it remains the default start for all job categories surveyed by YouGov, including those deemed best suited to AI. If your 2026 roadmap was doing SEO, schema, or technical clarity and chose “AI visibility,” this data means reversing that.
Second, create content that lasts a click, not just a quote. Since 22% of AI searchers click through to the links provided and only 17% stop at an AI answer, citations within an AI answer are not the bottom line. Layout pages so that anyone who clicks on an AI response lands on something more detailed, more current, and more clearly available than what was just summarized by the chatbot. That’s what turns a quote into a session.
Third, treat the status of “legitimate source” as a trusted asset, not the beauty of the product. Clear source links and a legitimate source framework are the two attributes that sway AI searchers the most, at 16% and 15%, respectively. That means visible underlines, date revisions, method sections, and structured data that make it clear that your page is the primary source, not a summary of it. Do it for an audience you can’t deliver, meaning people trust AI-assisted answers enough to check the receipt. Don’t waste budget trying to design a signal of trust for the 49% who say nothing will change their mind. That battle cannot be won with a UX tweak.
My Take
The SEO industry spent the first half of 2026 treating AI assistants as a competitive channel to defend against. This report says the opposite is closer to the truth. AI search in the US is small, focused on people who already search regularly, and structurally dependent on the same validation logic that has been driving traffic back to primary sources. The opportunity is not to win the quote war; making sure that if someone is going to look for a receipt, as Fantino says, your site is where they find it.
Additional resources:
Featured image: Master1305/Shutterstock



