NextFin News - Americans are not rejecting artificial intelligence outright. They are saying, with unusual clarity, that they do not yet trust it to make life better. In a new Pew Research Center study published June 17, only 16% of U.S. adults said AI will have a positive impact on society, while 40% said its impact will be negative and 31% called it mixed. That is not a trivial split. It means the country is moving into a phase where AI adoption is rising even as public permission for the technology remains weak.
The same study found that about half of U.S. adults now use chatbots and nearly a quarter say they use AI daily. Pew said that usage is up 16% from 2024, when just a third of adults reported using some kind of artificial intelligence tool. The contradiction is the story: Americans are increasingly exposed to AI through chatbots, smart speakers, smart doorbells, robot vacuums and thermostats, but exposure is not translating into optimism.
That matters because public sentiment is no longer an abstract backdrop to the AI boom. It is becoming one of the constraints on how quickly companies can roll out products, how regulators frame the issue and how much latitude politicians have to argue that the industry should self-police. Pew’s numbers suggest the mood has shifted from curiosity to caution.
One reason the findings matter is that AI is now operating in two directions at once. On one side, it is becoming more normal in daily life, with 49% of adults using chatbots and AI features moving into common consumer devices. On the other, the public is not warming up emotionally to the technology in the same way. That divergence can coexist for a while, but it usually becomes more expensive to manage as the technology spreads into work, media, education and household devices.
The survey also captures a deeper structural problem for the industry: the public appears to want the utility but not the institutions behind it. Pew found that 67% of adults have little or no confidence in the U.S. government to regulate AI effectively, and 60% are not confident companies will develop and use AI responsibly. Those are not niche concerns. They are the kind of numbers that shape how easily a new technology can win social license.
That combination of utility and skepticism is what makes the current AI cycle unusually fragile. The market can keep monetizing usage. The policy debate can keep escalating. But broad trust is not following the same curve as product adoption, and that gap is now the central fact in the story.
Negative Views Now Outnumber Positive Ones
The most important result in the survey is not just that only 16% see a positive impact. It is that the negative camp is much larger, and the middle is still substantial. Forty percent of adults said AI will have a negative effect on society, while 31% said it will have both positive and negative effects. That leaves only a small optimistic minority at the top end. In practical terms, AI is now facing the kind of public skepticism that usually arrives after a technology has already become embedded in daily life, not before.
That helps explain why the current debate around AI feels different from the early chatbot enthusiasm of 2022 and 2023. Early adoption was driven by novelty and productivity promises. The current environment is shaped by a mix of job anxiety, synthetic media, lawsuits and a growing sense that the technology is arriving faster than the rules around it.
“Only 16% say it will be positive.”
That single number is the sharpest signal in the report because it captures the asymmetry in public opinion. Support is thin, but concern is broad enough to matter politically. When a technology gets this kind of reception, companies can still sell it, but they have to spend more effort defending it.
There is also a generational and experiential angle hiding inside the aggregate. Americans who use AI daily are not necessarily enthusiastic about its broader social effects. That is an important distinction. Usage can be driven by convenience, curiosity or workplace pressure, while social trust is shaped by a wider set of worries: whether the systems are accurate, whether they replace jobs, whether they distort information and whether the people deploying them are behaving responsibly. Pew’s split suggests that exposure alone is not enough to close that trust gap.
In that sense, the survey reads less like a verdict on a product and more like a warning about legitimacy. A technology does not need majority admiration to become profitable. It does need enough public tolerance to avoid becoming a permanent political target. On the current numbers, AI has usage momentum but a thin legitimacy cushion.
Usage Is Rising Even As Trust Lags
Pew’s usage figures show why AI remains so commercially powerful despite the skepticism. About 49% of U.S. adults use chatbots, and nearly a quarter say they use AI daily. The technology is also moving into devices many people already own. Pew found that about a third of adults have a smart speaker, while AI features are appearing in smart doorbells for 18% of adults, robot vacuums for 13% and smart thermostats for 11%.
That blend of utility and unease is exactly what makes AI such a difficult public-policy issue. Consumers are not rejecting the product category; they are rejecting the idea that the industry has earned broad trust. Pew found that 67% of adults have little or no confidence in the U.S. government to regulate AI effectively, and 60% are not confident companies will develop and use AI responsibly. Those are not fringe numbers. They indicate a public that wants the tools but doubts the institutions around them.
“67% of US adults have little to no confidence in the US government to effectively regulate AI.”
That regulatory gap matters because trust is often the hidden prerequisite for scale. A technology can spread quickly when it is seen as useful, but it becomes harder to normalize when most people think neither government nor industry is ready to manage its downsides. The result is a market that can grow while still generating backlash.
That backlash does not have to take the form of a single dramatic event. It can build through a series of smaller frictions: workplace disputes over automation, debates over educational use, confusion about synthetic media, and consumer discomfort with AI features being embedded into products by default. Pew’s result helps explain why these arguments keep recurring even as usage expands. The public is not just debating capability; it is debating consent.
For companies, that distinction is important. Features can be shipped faster than norms can be settled. But if the norms turn hostile, adoption becomes more expensive and more political. That is already visible in how AI products are marketed: firms increasingly emphasize safety, guardrails, provenance and human oversight, not just speed and intelligence.
Why The Skepticism May Stick
The survey also suggests this is not just a temporary reaction to headlines. AI has moved from a text-based novelty into a system that touches images, video, search, consumer devices and workplace tools. As the stakes rise, so do the perceived risks. The public is not reacting to one product failure or one bad quarter. It is reacting to a widening sense that AI is becoming harder to distinguish from the rest of the digital economy — and harder to control once it is embedded.
That may help explain why the middle ground is so large. Thirty-one percent of adults said AI will have both positive and negative effects, which is effectively a verdict of uncertainty. Many Americans appear to believe the technology could be useful, but not without real costs. That is a harder environment for evangelists than simple opposition would be, because it forces them to prove benefits continuously rather than rely on a one-time promise.
“40% said its impact will be negative and 31% called it mixed.”
For the AI industry, that is a warning sign. Consumer adoption can still rise. Corporate spending can still accelerate. But the social license is narrower than the business momentum. If the public keeps associating AI with job displacement, misinformation, surveillance or unreliable outputs, the political response will eventually harden, even if the products themselves keep improving.
That risk is amplified by how quickly AI has moved from a specialized tool to an ambient layer in everyday digital life. When a technology is confined to a narrow use case, people can compartmentalize it. Once it appears in search, messaging, office software, devices and public services, it becomes harder to keep the debate technical. It becomes civic.
The broader implication is that the next phase of the AI story may be less about model capability and more about institutional credibility. If government regulators remain weak in the public mind and companies keep asking users to accept AI features before trust has been earned, skepticism may stop being a passing mood and become the default setting.
What It Means For The Market
For now, the market consequence is less about an immediate demand shock than a longer-term governance burden. Companies building AI products may keep benefiting from strong usage trends, but they will face more scrutiny over product design, disclosures, safeguards and model behavior. Regulators, meanwhile, have a public opinion tailwind for tighter oversight even if they lack a strong record of competence.
That could shape everything from consumer product launches to enterprise procurement. Enterprises generally do not need the public to love a technology before they buy it, but they do need a stable legal and reputational backdrop. If skepticism toward AI continues to harden, procurement teams will have more reason to slow-roll deployments, ask for audit trails and demand proof that the efficiency gains outweigh the risks.
The same logic applies to policy. A public that doubts both the industry and the government is unlikely to grant either side much benefit of the doubt when a controversy hits. That makes the system prone to sharp swings: one failure can bring a wave of criticism, while one useful feature can still drive adoption. The result is not a clean yes-or-no verdict on AI, but a more volatile environment in which trust becomes the swing variable.
The central takeaway is that AI has not lost the race for adoption. It has lost, for now, the race for public trust. That gap matters because the most durable technology cycles are not built only on usage; they are built on acceptance. Until AI closes that gap, every new product launch will be judged not just on what it can do, but on whether Americans believe it should be doing it at all.
For investors, policymakers and companies alike, the next catalyst is not just another product release. It is whether AI systems can become ordinary enough to be useful and trusted enough to be unremarkable. Right now, the numbers say that remains an open question.
Explore more exclusive insights at nextfin.ai.
