Most Americans don’t need an economist to tell them what they already feel in their bones: artificial intelligence is going to make the wealthy wealthier and leave everyone else further behind. A new poll confirms that suspicion has hardened into consensus — cutting across party lines, income brackets, and education levels in ways that should alarm the companies and policymakers betting big on AI as an engine of broad prosperity.
The findings come from a Gizmodo report on polling conducted by Bentley University and Gallup, which surveyed more than 5,000 U.S. adults. The headline number is stark: 75% of respondents believe AI will increase wealth inequality. Only 5% said it would reduce the gap. The remainder expected no change or had no opinion.
That’s not a partisan split. It’s a national one.
Republicans, Democrats, and independents all arrived at roughly the same conclusion. So did respondents across income levels and educational attainment. The rare issue where Americans actually agree: AI’s economic spoils will flow upward.
The Bentley-Gallup Business in Society Report, which housed the survey, also found that a majority of Americans — 53% — believe businesses are the primary beneficiaries of AI adoption, while just 7% said workers benefit most. The pattern is consistent with a broader erosion of public trust in the idea that technological progress automatically distributes its gains widely. That compact, central to Silicon Valley’s self-image for decades, appears broken in the public mind.
And the timing matters. These numbers land as AI investment reaches fever pitch. Microsoft, Google, Amazon, and Meta have collectively committed hundreds of billions of dollars to AI infrastructure. Nvidia’s market capitalization has swelled past $3 trillion on the back of demand for AI chips. Venture capital is flooding into AI startups at rates not seen since the dot-com era. The gap between corporate enthusiasm and public skepticism has rarely been wider.
The concern isn’t abstract. Workers in customer service, content creation, data entry, coding, legal research, and logistics are already seeing AI tools absorb tasks that once required human labor. Some companies have been transparent about headcount reductions tied to automation. Others have been less forthcoming, framing AI as an “augmentation” tool even as hiring slows in the very roles the technology targets.
A January 2025 report from the International Monetary Fund estimated that roughly 40% of global employment is exposed to AI, with that figure climbing to 60% in advanced economies. The IMF warned that AI could deepen inequality both within and between nations, depending on how governments respond. The fund’s managing director, Kristalina Georgieva, said at the time that “in most scenarios, AI will likely worsen overall inequality” — a projection that lines up precisely with what the American public already suspects.
The Trust Deficit Between Silicon Valley and Main Street
What’s striking about the Bentley-Gallup data isn’t just the top-line pessimism. It’s the depth of it. When asked whether AI will help or harm job seekers, respondents were sharply negative. When asked whether AI will be used ethically by corporations, the skepticism was equally pronounced. Americans increasingly view AI not as a neutral technology but as an instrument of corporate power — one designed to concentrate gains among shareholders and executives while shifting risks onto workers and consumers.
This perception has real consequences. Public resistance can shape regulation, slow adoption in sensitive sectors, and create political pressure that constrains how aggressively companies deploy AI. Already, the European Union has moved ahead with the AI Act, imposing risk-based restrictions on certain applications. In the U.S., the regulatory picture remains fragmented, but state-level proposals targeting AI in hiring, insurance, and healthcare are multiplying. The White House has issued executive orders on AI safety, though enforcement mechanisms remain thin.
Labor unions, meanwhile, have seized on AI as an organizing issue. The Writers Guild of America and SAG-AFTRA both made AI protections central to their 2023 strikes. Dockworkers negotiated limits on port automation. Teachers’ unions have pushed back on AI-driven assessment tools. The common thread: workers don’t trust employers to implement AI in ways that protect jobs or wages.
And there’s historical reason for that distrust. The productivity gains from earlier waves of automation — robotics in manufacturing, software in white-collar work — disproportionately accrued to capital owners. Real wages for most American workers have barely budged in decades when adjusted for inflation, even as productivity climbed. The promise that technology would lift all boats has, for many, curdled into a punchline.
AI boosters counter that previous technological transitions ultimately created more jobs than they destroyed. The internet wiped out travel agents and video stores but spawned entirely new industries. They argue the same will happen with AI — that new categories of work will emerge, that productivity gains will lower prices and expand access to services, and that the transition period, while painful, will prove temporary.
Maybe. But the polling suggests Americans aren’t buying it. Not this time.
Part of the skepticism may stem from who’s making the promises. Trust in big business has declined steadily for years. A 2024 Gallup survey found that only 14% of Americans expressed “a great deal” or “quite a lot” of confidence in big business — near historic lows. When the same institutions pushing AI are the ones Americans trust least, reassurances ring hollow.
There’s also a visibility problem. The benefits of AI tend to accrue quietly — a slightly faster insurance claim, a marginally better product recommendation, an incremental efficiency gain in a supply chain. The costs are loud: a layoff announcement, a viral story about AI hallucinations producing dangerous medical advice, a deepfake used for fraud. The asymmetry of attention works against the technology’s reputation even when the aggregate effects might be positive.
Still, the 75% figure in the Bentley-Gallup poll deserves to be taken at face value. It doesn’t just reflect ignorance or Luddism. It reflects a rational assessment of how previous technologies were deployed, who captured the value, and what the incentive structures look like this time around. Corporate leaders who dismiss this as a messaging problem rather than a structural one are making a serious mistake.
The question isn’t whether AI will generate enormous economic value. It almost certainly will. The question is whether any meaningful share of that value will reach the people who need it most — or whether the technology will simply accelerate the same dynamics that have been hollowing out the American middle class for forty years.
Right now, the American public has made its bet. And it’s not an optimistic one.
WebProNews is an iEntry Publication