Voters in political battlegrounds overwhelmingly favor the creation of a new federal agency to monitor and regulate artificial intelligence and the implementation of a government-designed test to ensure that AI models are safe, according to a new large-scale survey.
The polling suggests that lawmakers will face increasing pressure in the coming months to sign on to legislation creating new regulatory frameworks for the fast-developing technology, even as some in the industry warn that such rules could slow development and put the U.S. at a disadvantage to China.
The AI regulatory proposals had at least two-thirds support among both Democrats and Republicans in nearly all the competitive states and districts included in the University of Maryland’s Program for Public Consultation survey.
“Our surveys are clearly showing that large majorities of Americans of all political stripes want government to regulate AI,” Program Director Steven Kull told the Washington Examiner. “They may come to that conclusion for different reasons—concerns about deepfakes, jobs, data centers, or the potential for AI to act autonomously—but they arrive at the same place. They want safeguards.”
The survey assessed 11 competitive states and 28 competitive congressional districts, and included almost 20,000 adults.
Nationally, 85% of Democrats and 79% of Republicans favored creating a federal agency devoted to monitoring and regulating AI.
The Great American AI Act, a new bipartisan AI regulatory bill, would codify the existing Center for AI Standards and Innovation as the agency with oversight over AI.
Democrats and Republicans also agreed, at 78% overall, on requiring new AI programs that make critical decisions to first pass a government-designed test to ensure they follow regulations, protect data privacy and security, and do not have unintended biases.
An executive order signed by President Donald Trump in June, for instance, would have had frontier AI models go through such tests, but on a voluntary basis.
“I can’t think of another issue where Republicans and Democrats agree by margins approaching 80 percent,” the director of Federal AI Governance and Technology Policy at the advocacy organization Public Citizen, J.B. Branch, told the Washington Examiner. “This survey is a reminder that lawmakers represent voters.”
Lawmakers who have backed AI guardrails, such as the survey respondents, have received support from bipartisan organizations and industry leaders.
Although Trump won the election promising to keep AI free from many kinds of regulations, officeholders are facing new demands to impose rules.
The safety group Public First Action, for instance, recently announced a $15 million project to support Republicans holding AI companies accountable. Anthropic CEO Dario Amodei, similarly, donated $1 million to the AI regulation super PAC Public First.
One agenda item with strong bipartisan support is requiring labels for deepfakes, hyper-realistic AI-generated media that can possibly spread defamatory lies and damage an individual’s reputation, and prohibiting their use for political campaigns.
Lawmakers are also especially interested in drafting rules regarding children’s safety and age verification.
HOUSE REACHES DEAL ON PORNOGRAPHY AGE VERIFICATION AND SOCIAL MEDIA PROTECTIONS FOR MINORS
The SAFE KIDS Act, for example, would require AI systems to implement child-safety-by-design. The House has also reached an agreement on the Kids Internet and Digital Safety Act, which includes pornography age verification and social media protections for minors.
Free-speech advocates and some lawmakers have warned that the provisions would create opportunities for unauthorized surveillance or censorship.