With Election Day less than one week away, a new poll finds Maine voters overwhelmingly favor safeguards on artificial intelligence, and most say a candidate who supports them would benefit politically.
The Tyson Group, a national public opinion research firm, surveyed likely general election voters in Maine and found 74 percent would rather see reasonable safeguards than prioritize a race for global AI leadership. 75 percent said AI should be developed more cautiously, compared with 6 percent who want it developed faster.
The poll, summarized in a memo, tested how much the issue matters in a competitive race.
“Maine general election voters are clear: they want AI regulated, they agree on what the rules should be, and they reward candidates who back them,” the memo said.
Seventy-seven percent said backing AI safeguards would help a Maine candidate, while just 7 percent said it would hurt. Voters were similarly clear about candidates who push back. Even when they cite innovation, 74 percent said they would be less likely to support a candidate who opposes safeguards.
Every protection the poll tested drew more than 80 percent support, with children's safety leading the list. Ninety-three percent back stronger parental controls over children's use of AI. Large majorities also want AI systems monitored before and after release.
The support crosses party lines. At least three-quarters of Republicans backed every protection, and 72 percent of voters who backed Donald Trump said they would be more likely to support a pro-safeguard candidate.
The most notable finding for candidates is that neither party has claimed the issue.
“A party that acts on AI regulation will gain a critical advantage in this key state. Democrats have the head start, but Republicans can close it: 72% of Trump voters say they are more likely to back a pro-safeguard candidate,” the memo reads.
That leaves the issue open in the final days of the campaign.
The survey was conducted in October with a margin of error of plus or minus 5.7 percentage points.
