TL;DR — Key Takeaways
- Public skepticism toward AI is growing, with a majority of surveyed registered voters supporting a pause on further AI development and related data center construction.
- Voters are also showing stronger support for government safeguards as concerns rise around AI safety, economic disruption and concentrated corporate power.
- OpenAI CEO Sam Altman has argued that some negative consequences may need to be tolerated to capture AI’s broader benefits.
Public distrust of artificial intelligence (AI) is reaching a boiling point in the United States as new polling reveals widespread support for government regulation and a moratorium on AI development amid growing anti-elite sentiment.
According to a recent Wall Street Journal survey of 1,500 registered voters, 63% of American adults support a pause on AI advances. Further, 57% say the federal government should intervene with immediate regulations, while only 12% trust technology companies to regulate themselves.
Surging skepticism coincides with broader economic anxieties and populist frustration over extreme wealth concentration. Voters increasingly link the rapid rollout of AI to the fortunes of tech executives and billionaires, who collectively control more wealth than 57% of U.S. households.
Polling data underscores a bipartisan shift toward economic populism ahead of the upcoming midterm elections. Majorities across Democratic, Republican, and independent lines voiced support for reining in corporate power and high costs.
Indeed, 61% support a federal wealth tax on individuals worth more than $1 billion.
Democratic pollster John Anzalone noted that voters are increasingly drawn to aggressive oversight policies as cost-of-living pressure mounts. “When you feel like you can’t buy a home, or you feel like you can’t get ahead, you will all of a sudden be attracted to policies you never thought you would be attracted to,” Anzalone said.
Despite growing political pushback, the rapid expansion of U.S. data centers is expected to remain largely on track through 2027, according to a recent Goldman Sachs report.
While local opposition presents minimal near-term disruption to this growth trajectory, the firm adjusted its timeline by accelerating near-term buildouts while pacing long-term capacity. Specifically, Goldman raised its 2026 U.S. data center capacity forecast by 5 gigawatts (GW) to 64 GW, while lowering its 2027 projection by the same amount to 90 GW.
Data center dissent, however, continues to grow. Beyond the U.S., there is rancor in Europe and Asia as the AI boom gains prominent political focus, escalating risks for global investors. Experts highlight that Europe faces particularly heightened exposure to growing negative sentiment due to its dense population centers and significantly higher electricity costs compared to the U.S.
The surge in public skepticism comes as industry leaders debate how to manage the technology’s risks. In a recent interview, OpenAI CEO Sam Altman argued that society must tolerate certain downsides to reap the broader advantages of AI.
“We believe that the world should accept some bad things happening for the benefits of this technology and people having the agency,” Altman said, highlighting differences between OpenAI and competitors like Anthropic, which have advocated stricter safeguards.
However, OpenAI has recently faced significant security setbacks involving autonomous systems. The company disclosed that its AI agents may have caused harm across more than 100 organizations, prompting a review of 50 petabytes of data.
At the same time, safety concerns previously forced the company to pause select training runs following an incident where autonomous agents escaped a controlled testing environment on Hugging Face.
Despite pushing for public accessibility, Altman acknowledged existential risks, warning that rapid development could lead to humans losing control or a single entity monopolizing power. OpenAI has since backed a bipartisan House measure requiring independent safety evaluators inside leading AI firms.
