WASHINGTON, D.C. — The AI Policy Institute released survey results on June 12, 2026, indicating broad bipartisan support among likely U.S. voters for mandatory safety reviews and government oversight of powerful artificial intelligence systems. The survey of 1,007 likely voters was conducted on June 11 and June 12.

A large majority of respondents expressed a desire for powerful AI systems to undergo formal safety reviews before their public release. Republicans demonstrated greater enthusiasm for government-led safety testing of AI models compared to Democrats, though more than half of voters across all political affiliations supported such testing. Two-thirds of respondents favored AI systems with guardrails over an outright ban.

Voters also showed a strong preference for banning AI entirely if the alternative was unregulated AI systems. Forty-seven percent of respondents would allow data centers if the AI systems being developed had safety and security standards, while 38 percent preferred to ban data centers completely. Over 60 percent of both Republican and Democratic respondents believed the federal government should establish clear safety standards for AI systems.

A large majority, over 80 percent, of all respondents indicated that AI companies should not develop AI systems smarter than humans until they can demonstrate control over these systems. Among Democratic participants, 84 percent shared this view. Peter Wildeford, director of policy at the AI Policy Network, commented on the survey findings, stating, "Americans want to do more on AI safety."

In a related development, Commerce Secretary Howard Lutnick directed Anthropic to take its two most powerful AI systems offline in mid-June 2026 due to national security concerns. A June 2026 Pew Research Center survey found that approximately two-thirds of Americans believe AI is progressing too quickly. In that same Pew survey, 74 percent of Democrats and 61 percent of Republicans expressed a lack of confidence in the government's ability to regulate AI.

Reflecting on an earlier period, 2024 Pew Research Center data showed 70 percent of Republican-aligned voters and 54 percent of Democratic-aligned voters lacked confidence in the government's ability to regulate AI. Wildeford said, "I think people in the White House who have been trying to push a no-rules-whatsoever perspective are out of step with the American people on the results of the AIPI poll."

Why It Matters

The survey results from the AI Policy Institute show a desire among U.S. voters for government involvement in regulating artificial intelligence, particularly concerning safety and control. This public sentiment for oversight occurs as the government has begun taking actions, such as the Commerce Secretary's directive to Anthropic, to address perceived risks. The findings suggest a potential mandate for policymakers to develop regulatory frameworks, reflecting public preference for controlled AI development over either unregulated proliferation or total prohibition.