College of Maryland survey discovered that assist for AI regulation crosses celebration traces
In sum – what it’s worthwhile to know:
–United on AI – Massive majorities of U.S. residents, no matter celebration affiliation, are cautious of unregulated synthetic intelligence and wish to see AI regulation of some type that limits or guides the expertise’s improvement.
–Concern about competitiveness – Whereas survey members have additionally expressed that they don’t want authorities AI regulation to stifle improvement, and are involved about China outpacing the U.S. in AI, they had been extra involved concerning the potential harms of AI not being addressed till after the very fact.
A newly launched survey from the College of Maryland has discovered that there’s broad assist throughout celebration traces for presidency regulation of synthetic intelligence.
The survey was fielded from July thirtieth by means of August seventh, 2025 with 1,202 adults nationwide, by the Program for Public Session (PPC) on the College of Maryland’s Faculty of Public Coverage.
“Clearly Individuals are significantly involved concerning the present and potential harms from AI,” stated Steven Kull, director of PPC. “And whereas the general public is cautious of presidency regulation, they’re clearly extra cautious of the unconstrained improvement and use of AI.”
The survey discovered that whereas majorities of 59-69% of these surveyed agreed that they didn’t wish to stifle AI innovation (and thought the non-public sector may transfer quicker than the federal government to handle dangers), considerably extra individuals (77-84%) wished to see accountable innovation, agreeing that it was higher to be preventive than reactive concerning the potential harms of AI.
“As an alternative of letting issues out into the world after which reacting primarily based on flaws, they have to be functioning to their finest capability beforehand,” one participant instructed PPC as a part of the survey on AI regulation.
The PPC survey discovered that bipartisan majorities proceed to favor concepts resembling authorities certification of AI to guage whether or not fashions violate laws, have safety vulnerabilities or make biased selections (79% assist general, with 84% assist amongst Republicans and 81% assist amongst Democrats), and permitting authorities audits of AI software program that’s already in use, with corporations being required to repair any issues (78% assist, together with 82% amongst Republicans and 78% amongst Democrats).
The PPC survey additionally discovered assist for requiring corporations to speak in confidence to the federal government how decision-making AI was skilled, if that info was requested. There was additionally a really excessive assist (80% general, with assist stronger amongst Republicans) for prohibition or clear labeling of totally fabricated deepfakes in political promoting.
The findings had been in step with a earlier survey carried out by PPC within the spring of 2024, when a survey of greater than 3,600 registered voters concluded that “very giant bipartisan majorities favor giving the federal authorities broad powers to manage synthetic intelligence.”
Different makes an attempt to take the temperature of U.S. residents on how or whether or not they suppose AI must be regulated have had related findings. In April of this 12 months, Pew Analysis surveyed each consultants and laypeople about AI and located that whereas consultants had been typically extra constructive concerning the potential impacts of AI, about six out of 10 U.S. adults, and 56% of AI consultants surveyed, had been extra involved concerning the U.S. authorities not going far sufficient in regulating AI, than they had been about it going too far in regulation.
Individually, a June 2025 survey of greater than 1,000 registered voters, commissioned by tech advocacy group TechNet and carried out by Morning Seek the advice of, discovered that almost all of had been involved about China outpacing the US in AI improvement, and most well-liked a nationwide AI regulatory framework versus state-dominated regulation of AI.
Nevertheless, the Trump administration lately launched a federal AI Motion Plan that takes a intentionally mild hand with any limitations on AI, and orders a repeal of present federal laws or reexamination of authorized selections which “unnecessarily hinder” AI improvement or deployment. It additionally focuses on worldwide export of American AI expertise, with the most important AI ecosystem seen as the most effective.
“Whoever has the most important AI ecosystem will set international AI requirements and reap broad financial and navy advantages. Similar to we gained the house race, it’s crucial that the US and its allies win this race,” authors Michael Kratsios, assistant to the president for science and expertise; David Sacks, particular advisor for AI and crypto, and Marco Rubio — recognized as assistant to the president for nationwide safety affairs, relatively than as secretary of state — within the introduction to the AI Motion Plan.
“AI is much too essential to smother in forms at this early stage, whether or not on the state or Federal degree,” the plan stated. “The Federal authorities shouldn’t permit AI-related Federal funding to be directed towards states with burdensome AI laws that waste these funds, however also needs to not intervene with states’ rights to go prudent legal guidelines that aren’t unduly restrictive to innovation.” Nevertheless, the plan additionally directs that company heads with federal funding for AI “restrict funding if the state’s AI regulatory regimes might hinder the effectiveness of that funding or award.”
The AI Motion Plan directs the Nationwide Institute for Requirements and Know-how (NIST) to revamp with AI danger administration framework to remove references to misinformation, DEI and local weather change. It outlines promotion of a “try-first” AI local weather for American companies, directs federal companies to prioritize AI talent improvement and funding in AI adoption, together with in next-generation manufacturing to “usher in a brand new industrial renaissance” and in AI-enabled science and knowledge units; and drive adoption of AI inside authorities, together with within the Division of Protection.
