Majorities of both Republicans and Democrats say increased federal oversight is needed for artificial intelligence, according to a new survey showing broad bipartisan support for stricter safeguards on rapidly advancing AI technology.
An Artificial Intelligence Policy Institute poll found 65% of likely voters support a formal government review process for the most advanced AI models before they are widely released, while just 19% oppose the idea.
Support crossed party lines, with 66% of Republicans, 64% of Democrats, and 65% of independents backing the proposal.
The survey underscores growing concern that increasingly powerful AI systems could pose risks to national security if left largely unchecked.
The June 10-11 poll of 1,007 likely voters found Americans overwhelmingly favor targeted safeguards over either unrestricted development or outright bans on AI.
Nearly 6 in 10 respondents, 59%, said they support government safety testing of advanced AI models before release, while 65% said mandatory reviews should be required rather than relying on voluntary company participation.
Only 22% preferred voluntary reviews.
National security emerged as a dominant concern throughout the survey.
Seventy-one percent of respondents said the government's top AI priority should be keeping powerful AI technology out of the hands of America's enemies while maintaining America's strategic advantage.
Just 17% said policymakers should prioritize giving AI developers greater freedom to innovate.
Likewise, 70% supported requiring AI companies to implement stronger cybersecurity protections to prevent foreign governments from obtaining proprietary AI technology, while 64% backed export controls restricting AI software and hardware from reaching China, Russia, and other adversarial nations.
The poll also found voters overwhelmingly favor regulatory guardrails over a hands-off approach.
Seventy-eight percent preferred requiring advanced AI developers to implement safety and security standards compared with only 8% who favored no additional regulation.
Another 66% preferred guardrails rather than temporarily banning development of AI systems more powerful than today's leading models.
Support extended to accountability measures as well.
Sixty-nine percent backed the bipartisan AI Whistleblower Protection Act, which would shield AI company employees who report safety or national security concerns from retaliation.
The findings come as lawmakers debate how to balance innovation with security amid fierce global competition, particularly with China.
The Hill noted that Congress and state governments are also grappling with the rapid expansion of AI data centers, which have sparked concerns over electricity demand, environmental impacts, and local infrastructure.
While Americans remain divided over data centers in their own communities, the survey found 47% support allowing new facilities if accompanied by AI safety guardrails, while 38% oppose additional data center construction altogether.
Overall, the results suggest voters increasingly favor a regulatory framework that protects national security and public safety while allowing the U.S. to remain the global leader in artificial intelligence.