Less than four years after ChatGPT burst onto the scene, artificial intelligence stands at a crossroads. Financial markets celebrate its potential with soaring valuations. Yet public sentiment has soured. Polls reveal deep skepticism. Concerns over job losses, soaring electricity bills and environmental strain dominate conversations from boardrooms to town halls.
A MarketScreener report from May 20 captured the shift. Seventy percent of Americans believe AI advances too quickly. Fears of mass layoffs and higher power costs fuel the discontent. Data centers devour land, water and electricity. Local opposition has led to a record number of project cancellations in the first quarter of 2026.
But. This backlash runs deeper than isolated gripes. It now shapes policy debates in Washington and state capitals. And it forces technology executives to confront a reality they once dismissed.
Executives at leading AI firms once spoke of inevitable progress. OpenAI CEO Sam Altman warned there would be “real pain” and that “whole categories of jobs” will disappear. JPMorgan Chase CEO Jamie Dimon echoed the point. AI “will eliminate jobs,” he said. “People should stop sticking their head in the sand.” Those comments, highlighted in a recent Council on Foreign Relations analysis, underscore the tension. Productivity gains at the macro level offer little comfort to workers facing displacement.
Public opinion data paints a stark picture. Only 26 percent of voters hold positive feelings toward AI. Forty-six percent view it negatively, according to an NBC News poll cited across multiple reports. A Pew Research Center survey found just 10 percent of Americans feel more excited than concerned. Negative views have climbed from 34 percent three years ago to over 50 percent now, per YouGov polling detailed in an Axios article published May 17.
The disdain crosses party lines. More than 70 percent of Americans say AI development moves too fast. Sixty-eight percent of Republicans and 77 percent of Democrats agree. Among young people ages 14 to 29, only 18 percent feel hopeful about the technology, a Gallup survey showed.
Energy demands compound the anxiety. Data centers require enormous power. They drive up rates for ordinary consumers. Communities resist new facilities. This local pushback has emerged as a binding constraint on expansion. Investment banks Morgan Stanley and Jefferies noted in recent notes that canceled projects sap investor confidence.
Environmental worries add another layer. AI infrastructure consumes water and land. It contributes to carbon emissions at a time when climate concerns remain high. These practical impacts turn abstract fears into tangible grievances. Residents see their utility bills rise while tech giants post record profits.
Political leaders have taken notice. State lawmakers introduced more than 1,200 AI-related bills in 2025. Proposals range from disclosure rules to safety testing and outright moratoriums on data centers. In March 2026 the White House released a national AI policy framework. It sought to preempt some state rules while backing targeted federal steps on child safety.
Yet pressure builds from both sides of the aisle. Senator Bernie Sanders and Representative Alexandria Ocasio-Cortez called for a national moratorium on data center construction until stronger regulations take hold. President Trump, who benefited from tech executive support in his 2024 campaign, signed an executive order centralizing regulation at the federal level. He later announced a “Ratepayer Protection Pact” telling big tech companies they must meet their own energy needs. Allies have urged stricter testing for the most powerful models before public release.
The dynamic creates risk for the United States. AI drives economic growth, technological edge and military advantage. Yet it remains broadly unpopular domestically. Only a small fraction of citizens express enthusiasm. This mismatch could undermine long-term competitiveness if political responses favor pauses over careful progress.
Some AI companies have responded with public relations efforts. They organize around a simple message. We come in peace. OpenAI published a set of policy proposals addressing job replacement and wealth concentration. Ideas include a four-day workweek and a public-wealth fund seeded by AI gains and distributed to citizens. The Wall Street Journal reported on these charm offensives in April. Protests, such as one outside Google DeepMind offices in London, signal that messaging alone may not suffice.
Experts question whether the industry grasps the scale of discontent. Rahul Vohra, CEO of Superhuman Mail, seemed unaware of the polling when asked about backlash. “We don’t really see that,” he said. Dr. Avriel Epps, an assistant professor at the University of California, Riverside, offered a sharper view. “What is not inevitable is that these technologies will be embedded in every aspect of our lives, become indispensable, or replace humans.” Nothing in the future is inevitable, she added. No single company or group decides the outcome.
Arun Bahl, CEO of Aloe, struck a similar note. “Some version of AI is inevitable … but we have choice. Is it the dystopian plot? Or can we have tools that humans trust?” His comments appear in the same Axios report.
Globally, attitudes differ. A Stanford HAI AI Index found 59 percent of respondents in 2025 expect AI to do more good than harm, up from 55 percent the prior year. In the United States the picture remains gloomier. Surveys from the Annenberg Public Policy Center, Gallup and others show majorities want stronger regulation. Eighty percent support maintaining rules on safety and data security even if it slows development.
Capability growth continues at a rapid clip. Leading models have doubled in power roughly every four months since 2024. Projections suggest they could be 250 times more capable by the 2028 election. Labs now withhold their most advanced systems from public release over misuse fears. Military applications, from rapid targeting to vulnerability detection, demonstrate both promise and peril.
These advances occur against a backdrop of eroding trust. Americans worry about privacy, surveillance potential and unequal gains. They expect government to prevent unsafe products. They reject unchecked monitoring of daily life. And they demand the United States not lose its edge to competitors such as China.
The coming years will test whether industry, policymakers and the public can align. Current near-zero federal regulation appears unsustainable. Draconian pauses risk ceding ground abroad. Thoughtful approaches must balance safety, broad economic sharing and strategic leadership.
Recent commentary reinforces the stakes. A May 27, 2025, Brookings Institution piece noted backlash could reverse gains if abuses proliferate. Gary Marcus has warned repeatedly on X and in his Substack that sentiment will intensify. Public pushback, once dismissed, now influences capital allocation and regulatory agendas.
AI’s unpopularity no longer registers as mere noise. It has become a market force, a political reality and a strategic vulnerability. Companies that ignore it court greater resistance. Leaders who address it directly may yet shape a version of the technology that earns broader acceptance. The window for such alignment narrows with each passing quarter. Real pain, as Altman noted, already looms for many. How society distributes that burden will define the next chapter.
WebProNews is an iEntry Publication