America is not rejecting AI, but it is rejecting the speed and secrecy around it. That turns trust into a real distribution cost for every startup trying to sell AI now.
The AI market has a problem that cannot be solved with a better demo. Most Americans now say the technology is moving too fast, and twice as many say they are pessimistic about its long-term impact on society as say they are optimistic.
That is not a small branding issue for founders. It is the early shape of a go-to-market tax. If a buyer believes AI is useful but unsafe, interesting but untrustworthy, powerful but poorly governed, the sales conversation becomes longer before the product even gets judged on merit.
According to YouGov's latest Economist/YouGov Poll, conducted May 9 to 11 and published May 12, 71% of Americans say AI development is moving too fast. Only 27% say the pace is about right, and just 2% say it is moving too slowly. On the broader question of society, 51% say they are pessimistic about AI's long-term effects, compared with 25% who are optimistic.
The split cuts across the political map. Democrats are the most likely to say AI is advancing too quickly, at 77%, but Republicans are not far behind at 68%. Independents sit in the middle at 69%. For an industry that often frames regulation as a partisan threat, this matters. Public anxiety around AI is not living neatly inside one party or one age group.
The awkward part for the AI industry is that Americans are using the tools anyway. A February 2026 Economist/YouGov poll found that 63% of Americans have used AI, but only 23% use it regularly. The same poll found that 58% say they trust AI little or not at all.
That is the real tension. AI has crossed into daily life before it has earned daily confidence. People are trying chatbots, AI search, workplace copilots and image tools, but many are doing so with one hand on the brakes. This is not the adoption curve of a product people simply love. It is the adoption curve of a product they feel they may have to understand before it changes their job, their child's classroom, their medical records or the price of electricity in their neighborhood.
Pew Research Center's March 2026 roundup points in the same direction. Half of U.S. adults say the increased use of AI in daily life makes them more concerned than excited, while only 10% say they are more excited than concerned. Pew also found that Americans are more open to AI in data-heavy uses, such as analysis and medical research, than in areas that feel personal, creative or relational.
That distinction is useful for startups. Consumers are not saying every AI product is equally threatening. They are drawing lines. They may accept AI that helps a doctor review scans, a finance team detect fraud or a small business summarize customer tickets. They are more cautious when AI appears to replace judgment, companionship, creativity or control.
The Trust Gap Becomes A Sales Problem
For founders, skepticism shows up in practical ways. A customer asks where the data goes. A school district wants proof that student work will not train a model. A hospital wants a human-in-the-loop workflow. A bank asks who is liable when the system gets something wrong. None of those questions are fringe concerns anymore. They are becoming standard procurement.
This will reward companies that treat transparency as product design, not compliance decoration. Clear opt-in flows, plain-language model disclosures, audit logs, data retention controls and visible human review may sound less exciting than model performance, but they can shorten the distance between interest and purchase.
That is especially true because the economic worry is concrete. YouGov found that 64% of Americans say it is unlikely AI will create economic gains that benefit everyone. Younger adults are more likely than older adults to believe AI could produce broad economic gains, but they are also more likely to worry that AI could replace jobs they or their families depend on. Among adults under 30, 60% are somewhat or very worried about that risk.
Lower-income households also feel the threat more sharply. YouGov found that 56% of Americans with family incomes below $50,000 are somewhat or very worried about AI replacing jobs they rely on, compared with 47% among higher-income Americans. That matters because AI is being marketed as productivity software, but many people hear a different message: fewer workers, fewer choices and fewer guarantees.
The companies that ignore that will spend more on persuasion. They will need more case studies, more legal review, more customer education and more executive reassurance. They may still grow, but growth will be slower and more expensive than the AI boom narrative suggests.
There is another path. Startups can make control part of the pitch. Let customers choose where AI is used. Show when a human is responsible. Give administrators the ability to disable sensitive functions. Build products that improve a worker's leverage before promising to remove the worker from the process.
Regulation may also become less of a burden than many founders assume. If public trust is low, clear rules can help serious companies separate themselves from reckless ones. A market with standards is not necessarily a market with less innovation. Sometimes it is the market where buyers finally feel comfortable signing the contract.
The next phase of AI will not be judged only by capability. It will be judged by consent, accountability and whether ordinary people believe the benefits are being shared. Founders who understand that early will not just sound more responsible. They will sell into the market that actually exists.
Also read: AI data centers are testing rural America's bargain with Big Tech • Tencent's revenue miss puts its AI payoff under pressure • Metis TechBio's Hong Kong debut shows AI biotech is back in demand
