Public concern about artificial intelligence is becoming a political force across North America and Europe, with roughly half of adults in six countries saying development should pause while governments work to reduce the risks. The survey, conducted by Public First from Sept. 13 to 15, questioned more than 2,000 people each in the United States, Canada, the United Kingdom, France, Spain and Germany. Its overall margin of error was plus or minus 2.2 percentage points, with larger margins for smaller groups. In every country surveyed, more people favored pausing AI development than continuing it solely to pursue potentially significant benefits. Support for a pause ranged from 45% in Germany to 51% in the United Kingdom. Support for continued development ranged from 30% to 40%. At least half of respondents in each country said there was a moderate or greater risk that AI could “destroy humanity.” That view was strongest in the United States, France, the United Kingdom and Canada, where it approached two-thirds of adults. The results also indicate that many people would accept their country falling behind in the global AI competition if doing so meant greater safety and accountability. Majorities in Canada and the United Kingdom, each at 53%, chose security over maintaining a technological lead. Respondents in several other countries also leaned toward caution. The public skepticism comes as governments debate how to regulate rapidly advancing systems while trying to build domestic computing capacity and attract investment. The European Union adopted the first comprehensive AI law in 2024, but has relatively few companies developing the most advanced models. European leaders are now attempting to balance safety measures with efforts to compete internationally. Activists have pushed for a slower pace as well. PauseAI demonstrators gathered outside the European Parliament in Brussels, where Belgian coordinator Didier Coeurnelle compared the risks to boarding a plane with a one-in-ten chance of exploding. Warnings from prominent technology executives have helped bring the issue into mainstream politics. Anthropic chief executive Dario Amodei, OpenAI chief executive Sam Altman and Elon Musk have all raised concerns about the consequences of advanced AI, while other industry figures have rejected calls for stronger government controls. Neil Chilson, head of AI policy at the Abundance Institute, described the findings as a familiar technology-driven panic: relatively few present-day harms paired with widespread fear about future ones. He noted that at least one-third of respondents in every country except Spain said they rarely or never use AI. “Most people haven't felt AI's benefits but they've heard many scary stories about its future,” Chilson said. “For now, fear is outrunning experience. But experience will catch up.” The survey found that voters generally viewed warnings from AI researchers as sincere rather than a marketing strategy, though political differences were more pronounced in the United States. Among Americans who voted for Donald Trump in 2024, 44% favored pausing advanced AI development and 40% preferred moving ahead. Trump voters were also more likely than Kamala Harris voters to view warnings from AI employees as promotional tactics and to oppose letting other countries gain an advantage. The divide has not prevented action at the state level. Republican and Democratic state officials have pursued tighter controls despite resistance from the federal government. In California, Gov. Gavin Newsom signed an executive order advancing AI oversight standards and directed state agencies to study whether the most advanced models should be required to have a shutdown mechanism. The political stakes are likely to grow as AI affects employment, public services, elections and energy use. In the survey, respondents in each country were more likely to believe AI would destroy jobs than create them. For policymakers, the findings suggest that demands for guardrails are no longer confined to technology critics or one part of the political spectrum.