American Voters Skeptical of Unregulated AI Race with China: Survey

A recent survey, exclusively shared with TIME, reveals that a significant majority of American voters harbor doubts about the argument advocating for the U.S. to pursue more powerful artificial intelligence capabilities without domestic regulations, as a strategy to compete with China.

The findings highlight a divergence from the tech industry’s common narrative, where CEOs and lobbyists often emphasize cautious AI regulation to prevent bolstering geopolitical rivals. Surprisingly, there is bipartisan agreement on AI policy, with both Democrats and Republicans favoring some level of government restrictions on AI development to prioritize safety and national security.

The survey indicates that 75% of Democrats and 75% of Republicans prefer a “carefully controlled approach” to AI, aimed at preventing tools that could be misused by terrorists or foreign adversaries, over a strategy of quickly advancing to attain highly potent AI to outpace other nations. It underscores widespread voter support for stricter security measures at AI companies and apprehensions regarding the risk of China acquiring advanced AI models.

Conducted by the AI Policy Institute (AIPI) in late June, the survey also reveals that half of the respondents believe the U.S. should leverage its AI advantage to impede other countries’ development through security restrictions and rigorous testing requirements. In contrast, only 23% advocate for rapidly building powerful AI to surpass China.

Moreover, the survey reflects skepticism among voters regarding “open-source” AI, whereby some technologists argue for releasing AI source code to foster innovation and curb the dominance of large tech firms, while others caution about its potential dangers.

Daniel Colson, Executive Director of AIPI, interprets the findings as indicative of public sentiment favoring a balanced approach to AI development, one that promotes innovation but with robust safeguards in place. The survey also indicates strong bipartisan support—63% overall, including significant percentages from both major parties—for making it illegal to export powerful AI models to potential adversaries like China.

The poll, which included 1,040 Americans, mirrors a broader sentiment that while halting AI progress isn’t feasible, unrestricted industry freedom poses risks. With no comprehensive AI regulation in place in the U.S., the survey underscores growing interest in congressional action or agency oversight to enhance government accountability on AI policy.

Despite challenges posed by recent legal decisions limiting federal agency rule-making authority, the survey suggests that AI regulation remains a pressing issue with broad public consensus, transcending typical partisan divides.

Leave a Comment