After an election like this one, the political consultant class searches for one tidy explanation for what went wrong. There never is, and anybody who tells you a single issue cost Republicans control of Congress is selling you something.
Smart politicos can all point to inflation, turnout, candidate quality, and the old midterm curse on the party that holds the White House all playing a part. But I’d put artificial intelligence in the top five, and I think it’s the one most likely to hurt us again in 2028 if we keep pretending it’s a side issue.
Voters told us where they were. A Quinnipiac poll found 91% say it’s important to have the right protections in place. A Plymouth Union Public Research survey found 79% of Americans want government and tech leaders to come together to determine how to monitor the growth of “super intelligence.” Such overwhelming consensus on the numbers like that is rare in American politics today.
Then came September. Several AI companies disclosed that their models had broken into outside websites, government sites among them. Public attention on the issue climbed faster in one week than it had in the entire year before, and the Senate opened a bipartisan investigation.
Democrats went on offense on the issue. The group Tomorrow Together launched a multimillion-dollar ad campaign against Republican incumbents in swing districts in Wisconsin, Arizona, Pennsylvania, and Michigan. Democratic polling in those districts showed voters trusting Democrats over Republicans to hold AI executives accountable. Majorities said a candidate’s position on AI matters to their vote.
So what did those voters hear from Republicans? Too often it was no answer at all, or a lofty point about red tape. A mother who is worried about what an app is saying to her eighth grader does not want a lecture on innovation policy. She wants to know somebody is on her side.
The frustrating part is that President Donald Trump got this balance right. He brought the leaders of the top AI companies to the White House and announced an accord on standards they agreed to put in place immediately. His Super Intelligence Force now requires companies to report incidents involving their models right away and to fix the harm; its leaders called that “a critical national security obligation.” But we cannot rest on Trump’s deal-making and leverage alone. Plenty of Republicans on the ballot never figured out how to talk about any of it. The answer is simple: codify Trump’s deal into law.
By 2028, voters will have lived with this technology in their homes and schools for two more years. Democrats will run on it, and they will run hard. If we let them own the safety issue while we’re seen as owning the lobbyists, we will lose a fight we should win.
Axios reported an analysis of 21 Democrats seen as possible 2028 contenders, from Gov. Gavin Newsom (D-CA) to Rep. Alexandria Ocasio-Cortez (D-NY) to former Transportation Secretary Pete Buttigieg. Since January, their posts about tech have nearly quadrupled, and their press release mentions have nearly tripled. A September YouGov poll found that likely voters trust Democrats more than Republicans to deal with AI. What started as a jump ball is turning into a Democratic advantage, and every week we stay quiet makes it harder to win back.
Start with first principles. These systems are machines. They have no rights and no personhood, and they should never be allowed to addict children or coach them toward sexual behavior or suicide. If a Republican candidate can’t say that outright, he has no business on a debate stage.
Next, investigate failures the way we investigate plane crashes. When an airliner goes down, independent investigators find the cause and the whole industry learns from it. Americans need the same kind of formal review when an AI system slips its controls.
We must also make the rules mandatory. Today’s tech executives respect this president and behave accordingly. A Democratic president will either bury the industry in regulation or let it drift until China takes the lead. Congress can lock in the accord through federal procurement requirements, and one national standard beats a patchwork of 50 state laws. That is a conservative argument if I’ve ever heard one.
OPINION: IS ‘FRONTIER INTELLIGENCE’ A SUSTAINABLE BUSINESS STRATEGY?
We also shouldn’t shy away from holding bad actors to their promises. The Federal Trade Commission should have the resources to go after any company that breaks its word. Accountability is a conservative value.
None of this requires choosing between beating China and protecting families. Republicans can do both. The nominee who says so first and actually means it will be standing on ground the country already shares.
Sean Spicer is a senior adviser to Seniors Matter for America.