Americans Split on AI Safety vs. China Competition, Poll Shows

Competing Visions for AI Development

Americans hold sharply conflicting views on how aggressively the United States should pursue advanced artificial intelligence development, according to new polling data that reflects an emerging partisan fault line in Congress.

While 63% of respondents in a recent POLITICO survey acknowledged AI poses genuine existential risks to humanity, they remain divided on the appropriate response. A plurality of 43% believe the U.S. should accept falling behind other nations if necessary to develop AI more responsibly and securely. Meanwhile, 37% argue America must maintain its lead even if doing so requires accepting greater risks.

The partisan breakdown is stark: voters who supported President Donald Trump in 2024 favor the competition-first approach by a margin of 50% to 38%, while those who backed Vice President Kamala Harris take the opposite position, prioritizing safety over speed.

Why It Matters

This polling data captures a fundamental tension that will shape U.S. AI policy for years to come. As Congress begins crafting legislation to govern frontier AI systems, lawmakers face pressure to balance two competing imperatives: preventing potential catastrophic outcomes from advanced AI while ensuring China doesn't establish technological dominance. The partisan split suggests bipartisan AI legislation will require difficult compromises on the core question of how much risk is acceptable in pursuit of global leadership.

Congressional Battle Lines Form

The public divide mirrors an increasingly visible split on Capitol Hill. Republican lawmakers interviewed this week emphasized the competitive dimension, warning that any pause in AI development could allow China to surpass American capabilities.

Rep. Brett Guthrie (R-Ky.), who chairs the House committee overseeing the technology sector, framed the issue in zero-sum terms at the POLITICO Decoded Summit on Wednesday. "We absolutely can't surrender," Guthrie said. "And I think that's what it would be if we stopped AI development."

Democratic lawmakers acknowledge the competitive landscape but resist the argument that competition should override safety considerations. Rep. Greg Casar (D-Texas), chair of the Congressional Progressive Caucus, offered a contrasting framework at the same summit: "The AI race is not an arms race that you win. It's an arms race that we need to end, and we need to protect the American people."

Casar has partnered with Sen. Bernie Sanders (I-Vt.) on legislation that would ban "superintelligence" development altogether—a proposal that gained renewed attention following warnings from a former Anthropic researcher last week about AI's potential to "kill us all."

Implications for Policy

The polling suggests any bipartisan AI legislation will need to thread a narrow needle. With AI safety moving to the center of congressional debate but legislation still in early stages, the competition-versus-caution question stands as a major obstacle to consensus.

The POLITICO Poll was conducted by Public First from September 13-15, 2026, surveying 2,064 U.S. adults online with a margin of error of ±2.2 percentage points. These findings were first reported by POLITICO.