76% of Americans Don't Trust AI Answers. Buyers Are Using Them to Build Your Shortlist Anyway.
A new Quinnipiac poll finds 76% of Americans distrust AI outputs even as fewer say they have never used it. For B2B brands, the paradox is the strategy: buyers use AI to build shortlists, then validate through earned media your brand may not have.
A Quinnipiac University poll of nearly 1,400 Americans released March 30 found that 76% think they can trust AI-generated information only some of the time or hardly ever, while only 27% volunteered that they have never used AI tools, down from 33% in April 2025 (Quinnipiac University, March 2026). Buyers are adopting AI research at record rates while openly admitting they do not trust what it tells them. The gap between adoption and trust is not a contradiction. It is the mechanism that determines which brands survive the shortlist.
The paradox is the buying process
The obvious reading is that trust will catch up to usage. The data says the opposite is happening. Usage is accelerating while trust is falling.
Forrester's Buyers' Journey Survey, 2025, of nearly 18,000 global B2B buyers found that 94% now use AI during their buying process (Forrester, January 2026). But the same survey found buyers increasingly validate AI outputs against trusted external sources because AI tools "often deliver incomplete or unreliable information, creating mistrust." G2's 2026 Buyer Behavior Report, a survey of more than a thousand B2B software buyers, found eight in ten had sourced a software recommendation from an AI chatbot in the last two years.
The buyer workflow now operates in two distinct stages. The AI session generates the shortlist. The validation session tests whether the shortlist holds up. The shortlist is built on what AI models know about your brand. The validation is built on what independent publications have said about you.
| Stage | What the buyer does | Where your brand needs to exist |
|---|---|---|
| AI shortlist generation | Asks ChatGPT, Perplexity, or Copilot to compare vendors in the category | In the training data and indexed sources AI models draw from |
| Trust validation | Checks the AI recommendation against peers, analysts, and editorial coverage | In Tier 1 publications buyers already trust as humans |
| Final decision | Contacts 1-2 vendors the shortlist and validation both confirmed | On both lists, not just one |
The pre-contact favorite wins 80% of deals, according to the Forrester data. That favorite is now selected during an AI research session the vendor never sees. Then validated through editorial coverage the vendor may never have earned.
The 76% distrust number is the leverage point
Here is where most teams misread the Quinnipiac data. They see 76% distrust and conclude AI search does not matter yet. That is backward.
If buyers distrust AI the way the 76% of Americans in this poll do, most of them are looking for reasons to disconfirm what AI told them. The AI answer is the hypothesis. The validation step is the test. If your brand passes both, you win the deal. If you pass the AI shortlist but fail validation because your earned media is thin, you lose to whoever the buyer finds in Forbes or TechCrunch when they go looking for confirmation.
In Opollo's analysis of 312 IT and technology brands in its client portfolio (Q3 2024 to Q1 2025), AI-referred visits converted at 14.2% versus 2.8% for Google organic. One plausible reason for that premium is that the AI already pre-qualified the buyer before they arrived. But the Quinnipiac data adds a condition: the conversion only holds if the buyer validates the AI recommendation through independent sources before acting on it.
Ahrefs' study of 75,000 brands found brand web mentions correlate at 0.664 with AI Overview visibility, stronger than backlinks at 0.218 (Ahrefs, May 2025). In that sample, the brands mentioned most across the web were the most visible in AI Overviews. Our hypothesis is that the brands with the most earned coverage in trusted publications are also the ones buyers find when they validate. Both layers may point at the same asset: earned media in publications that carry independent credibility.
What the validation step actually looks like
The Quinnipiac poll found 51% of Americans now use AI for research, up from 37% in April 2025. But 80% are either very concerned or somewhat concerned about AI. The behavioral split is striking: people use AI precisely because it is fast and comprehensive, then immediately distrust it because they know it hallucinates, omits, and oversimplifies.
For B2B buyers, the validation step is not casual browsing. Forrester found that buying groups now average 13 internal stakeholders and 9 external influences per decision. When a procurement lead gets a shortlist from Perplexity, the next move is checking whether those brands have been covered in the publications the buying committee trusts. If your brand shows up in the AI answer but has no independent coverage to back it up, the validation step eliminates you.
Muck Rack's December 2025 report classified 82% of its sampled cited links as earned media, its broad category for sources a brand neither owns nor pays for (report). Its May 2026 edition reported 84% (primary report). These are AI citation samples, not observations of human validation. Our hypothesis is that useful coverage may serve both; their source overlap was not measured here.
Why distrust does not slow adoption
The Quinnipiac data showed only 6% of Americans are "very excited" about AI while 62% are not excited. Yet only about a quarter say they have never used it. This is not irrational. It is what happens when a tool delivers speed and breadth that nothing else matches, regardless of whether the user trusts each individual output.
The B2B buyer does not need to trust the AI answer. They need a starting point. The AI compresses what used to take two weeks of Google searches, review site browsing, and peer conversations into a 90-second session. The buyer gets a working shortlist in the time it takes to drink coffee. Then they spend the next three days validating it through the same trusted channels they always used: analysts, editorial coverage, peer recommendations.
Our recommendation is to build material useful to AI-assisted research and human validation, then measure each audience separately. Credible earned coverage is one priority, not a proven necessary condition for entering an AI shortlist or surviving every buyer decision.
FAQ
Does low AI trust mean brands should ignore AI search visibility? No. In the Quinnipiac data, the share of Americans who say they have never used AI tools fell by six points since April 2025, while only about one in five trust AI-generated information most or almost all of the time, largely unchanged from April 2025. Buyers use AI for speed, then validate through independent editorial coverage. We recommend useful material for both audiences and measuring which sources each actually uses; the studies here do not establish earned coverage as a mandatory entry condition.
What determines whether a brand survives the buyer's validation step? Earned media coverage in authoritative publications. Forrester's 2026 survey found buyers validate AI outputs against "trusted external sources." Ahrefs found brand web mentions correlate more strongly with brand visibility in Google AI Overviews than backlinks (0.664 vs. 0.218). Our hypothesis is that the validation and the AI recommendation draw from a similar source pool: independent editorial coverage; neither Forrester nor Ahrefs tested that.
How does the AI trust paradox affect B2B pipeline specifically? The pre-contact favorite wins 80% of deals (Forrester 2026). That favorite is now determined during AI research sessions where 94% of B2B buyers use AI tools. Forrester's survey measures buyer behavior; it does not show which brands AI answers cite, or that being cited produces pipeline. And Quinnipiac's poll of US adults found 76% trusting AI-generated information only some of the time or hardly ever; a buyer who doubts the answer seeks confirmation. A brand on the AI shortlist without earned media to confirm the recommendation loses to a competitor the buyer can verify.
The paradox is not going away. Adoption keeps climbing. Trust is not keeping up. The gap between the two is where your brand either gets confirmed or gets replaced. The brands that own both layers, the AI shortlist and the validation coverage, win the deal before the sales team knows the buyer exists.
Our strategy is to earn useful coverage that can serve machine research and human validation. A placement may be cited by an AI engine and found by a buyer, but the studies here do not establish identical source pools or guarantee either event. Machine Relations, coined by Jaxon Parrott, is the framework we use to build and measure that presence.
How visible is your brand to the AI agents building shortlists and the humans validating them? The AuthorityTech visibility audit shows exactly where your brand appears in AI answers and where the validation coverage is thin.
Correction, November 11, 2026
This page linked Quinnipiac University's findings on AI use and trust to a different Quinnipiac release, an October 2025 Pennsylvania poll, and dated the AI poll March 31; the national AI poll was released March 30, 2026, and the page now links it. It also said trust in AI fell, and gave a trust share and an adoption share that the poll does not state. Quinnipiac found that about one in five think they can trust AI-generated information most or almost all of the time, largely unchanged from its April 2025 poll, and that the share who volunteered they have never used AI tools fell from about a third to about a quarter; the FAQ, the description and the close now say that. The poll surveyed American adults, not buyers, so the headline and the FAQ now say Americans where they give its share.