There was a time, not long ago, when artificial intelligence carried the sheen of promise. Tech executives spoke in rapturous tones about productivity gains and creative liberation. Venture capital poured in at historic rates. The public, for its part, seemed cautiously intrigued.
That time appears to be over.
A new poll from the Pew Research Center has found that negative views of AI among Americans have reached a striking high, with the technology now ranking among the most disliked institutions and concepts in American life — more reviled than the IRS, more scorned than major banks, and viewed more unfavorably than social media companies. As Futurism reported, the data paints a picture of a public that hasn’t just grown skeptical of AI. It has grown hostile.
According to Pew’s findings, 53 percent of Americans now hold an unfavorable view of AI, while only 17 percent view it favorably. That net favorability score of negative 36 puts AI in remarkably grim company. The IRS, long a reliable punching bag in American public opinion, manages a better net score. So does Immigration and Customs Enforcement. So do the major tech companies themselves, even after years of privacy scandals and antitrust scrutiny.
Let that sink in. A technology that Silicon Valley has marketed as the most transformative force since the internet is now less popular than the agency that audits your taxes.
The Collapse of Public Trust Didn’t Happen Overnight
The erosion has been building for at least two years, accelerating sharply since ChatGPT’s public debut in late 2022. Pew’s tracking data shows that favorable views of AI have dropped steadily — from 25 percent in early 2024 to the current 17 percent. Unfavorable views have climbed in near-perfect inverse proportion. The trend line isn’t ambiguous. It’s a straight downward slide.
What’s driving this? Several forces appear to be converging simultaneously. Job displacement fears are the most obvious. Every week brings fresh headlines about layoffs tied to automation, companies replacing customer service teams with chatbots, and executives openly discussing plans to reduce headcount through AI adoption. For millions of workers, AI isn’t an abstraction. It’s a direct threat to their paycheck.
But the anxiety runs deeper than employment. Americans are increasingly concerned about AI-generated misinformation, particularly heading into election cycles. They’re worried about surveillance. They’re uncomfortable with the opacity of systems that make consequential decisions about loan approvals, medical diagnoses, and criminal sentencing. And they’re watching the technology’s most prominent champions — figures like Sam Altman and Elon Musk — engage in public feuds, corporate power struggles, and what often looks like reckless deployment of half-finished products.
Trust, once lost, is extraordinarily difficult to rebuild. The AI industry is learning this lesson in real time.
The demographic breakdown in Pew’s data reveals additional complexity. Younger Americans, who might be expected to embrace new technology most readily, are among the most skeptical. Workers without college degrees — those most vulnerable to automation — express the highest levels of negativity. And the partisan divide that characterizes so much of American opinion extends here too, though notably, dislike of AI bridges the political spectrum more than most issues do. Republicans and Democrats don’t agree on much in 2025. They increasingly agree that AI makes them uneasy.
This bipartisan discomfort matters enormously for the industry’s regulatory future. When both sides of the aisle share a negative view, legislation becomes far more likely. And indeed, AI regulation has gained momentum at both the state and federal level in recent months. The European Union’s AI Act is already being implemented. In the United States, a patchwork of state-level proposals is advancing, many of them focused on transparency requirements and restrictions on AI use in hiring and law enforcement.
Silicon Valley’s Messaging Problem Is Now a Business Problem
For the major AI companies, this polling data isn’t just a public relations headache. It’s a potential commercial crisis. Consumer-facing AI products depend on adoption, and adoption depends on a baseline level of trust. If more than half the country views your core technology unfavorably, the growth projections that justify hundred-billion-dollar valuations start to look shaky.
Consider the specific comparison to social media companies. Meta, X, and TikTok have all weathered years of public backlash over content moderation failures, teen mental health impacts, and data privacy violations. Yet even these companies manage better favorability numbers than AI as a category. That’s a remarkable indictment. Social media has been the tech industry’s villain for the better part of a decade, and AI has managed to leapfrog it in public contempt.
The industry’s response so far has been largely tone-deaf. Tech leaders continue to emphasize capability — what AI can do — while largely ignoring what the public actually cares about: what AI will do to them. Press events showcase ever-more-powerful models. Earnings calls trumpet efficiency gains. But the fundamental question most Americans are asking — “Will this technology make my life better or worse?” — goes largely unanswered by anyone with credibility.
And credibility is the core issue. The AI industry has made sweeping promises before. Self-driving cars were supposed to be ubiquitous by 2020. Virtual assistants were going to manage our lives. Each unfulfilled promise erodes the foundation for the next one. Now, when AI companies promise that their technology will create more jobs than it destroys, or that safety measures will prevent misuse, the public has ample reason to be skeptical.
Some companies are beginning to recognize the problem. Microsoft has invested heavily in what it calls “responsible AI” initiatives. Google has published frameworks for AI ethics. OpenAI has made public commitments to safety research. But these efforts often feel performative — corporate governance theater designed to forestall regulation rather than genuinely address public concerns. The Pew numbers suggest the public sees through it.
There’s a historical parallel worth examining. Nuclear energy, in its early decades, enjoyed broad public support as a clean, powerful alternative to fossil fuels. Then came Three Mile Island, Chernobyl, and Fukushima. Public opinion cratered, and it took decades — and a climate crisis — to begin recovering. AI hasn’t had its Chernobyl moment. But the steady accumulation of smaller failures, controversies, and broken promises may be producing the same result through a thousand cuts rather than a single catastrophe.
The business implications extend beyond consumer products. Enterprise AI adoption, which accounts for the bulk of industry revenue, also depends on public sentiment. Companies deploying AI in customer-facing roles risk backlash if their customers view the technology negatively. A bank that replaces human loan officers with AI systems may save money in the short term but lose customers who don’t trust algorithmic decision-making. A hospital that uses AI for diagnostics may face patient resistance. The savings from automation can be quickly offset by the costs of lost trust.
Wall Street, for its part, has been slow to price in this sentiment risk. AI stocks have surged on the strength of revenue growth and the assumption of continued adoption. But if public opinion continues deteriorating at the current rate, the adoption curve that underpins those valuations could flatten or even reverse. Smart investors are watching these polls as closely as they watch quarterly earnings.
So where does the industry go from here? The honest answer is that no one knows. The technology itself continues advancing rapidly, and there are genuine use cases that deliver real value — drug discovery, climate modeling, accessibility tools for people with disabilities. But the gap between what AI can accomplish in controlled settings and how it’s perceived by the general public is widening, not narrowing.
Closing that gap will require something the tech industry has historically been bad at: humility. It will require acknowledging that public fears aren’t irrational. It will require slowing down deployment in sensitive areas until safety and fairness can be credibly demonstrated. It will require engaging with critics rather than dismissing them as Luddites. And it will require accepting that some applications of AI — surveillance, autonomous weapons, deepfakes — may need to be restricted or banned outright, even if they’re technically possible and commercially attractive.
None of that is likely to happen voluntarily. Which is why the Pew data may ultimately matter most as a signal to regulators and lawmakers. When a majority of Americans view a technology unfavorably, democratic governments tend to respond. The question isn’t whether AI regulation is coming. It’s whether the industry will help shape it or have it imposed upon them.
For now, the numbers speak clearly. Americans have rendered their verdict on AI, and it isn’t the one Silicon Valley was hoping for. Not even close.
WebProNews is an iEntry Publication