Politico: How to pitch voters on the AI apocalypse
By AARON MAK | 01/14/2026 04:10 PM EST
Popular backlash against artificial intelligence could be a powerful force in the midterms.
A number of candidates successfully connected rising utility costs to data centers during last November’s elections — so much so that tech companies are now pouring billions of dollars into PR campaigns to change the narrative. Plus, AI-related job loss and mental health concerns continue to dismay the public.
AI is already having a profound effect on people’s lives, and many of those mobilizing to set guardrails on the technology are also concerned it could trigger a nuclear apocalypse or enslave humanity. While these theoretical dangers have longer-term implications for humanity, they can seem a bit sci-fi.
“Concerns about catastrophic risk — how terrorists might get access to it or misalignment risk — are ones that people are aware of,” said Brad Carson, a former Democratic congressman for Oklahoma who’s setting up super PACs to push for more AI regulations in the midterms.
“But they probably aren’t as salient child suicides, rising electricity bills — those kinds of concerns really speak to people.”
So, for AI regulation advocates who want to animate voters, is it shrewder to just focus on harms currently arising from the technology, or should the distant prospect of doomsday be an election issue as well?
According to advocates who talked with DFD, there’s room for both: Campaigning on more salient kitchen-table AI issues can lead to more abstract considerations about the course of human civilization.
“I want to connect these issues as much as possible because I think they really are connected — the longer term existential and societal risks, and the immediate impacts that people feel and see on their lives and communities,” said David Krueger, an assistant professor at the University of Montreal who studies AI and recently founded Evitable, an activist nonprofit that’s seeking to make the midterms a referendum on the potentially catastrophic perils of frontier AI development.
Krueger noted the first step will be to get voters to think about AI “not as an isolated narrow issue of misinformation or child safety, or labor impact or environmental impact.” He added, “These are all coming from the same source and we’re not going to do a good job responding to this if we’re just trying to play whack-a-mole with all of the problems that are coming out of AI.”
Once it’s established as a systemic concern, then voters can question the whole project of AI and the existential risks that accompany its development.
Carson wants voters to link quotidian struggles with the bigger picture dilemmas. “Electricity is the gateway drug of AI awareness,” he told DFD. “When you see your electricity bill goes up, suddenly your political consciousness is raised: what is happening with these data centers? What are they being used for?”
Midterm messaging could help voters make this connection if it’s crafted with some media savvy. “If you’re talking about 30-second ads [...] you have to talk about what’s most relevant — child safety, electricity bills — things that are obviously on the top of people’s minds today,” Carson said. He suggested that such ads may prompt voters to explore the issue in longer formats like podcasts, where AI regulation advocates have more space to address speculative concerns about future disasters.
Carson and Krueger note that some of the auditing policies they’ll be promoting in the midterms will help to spot harms that models could unleash now and in the future, which is another reason it makes sense to take an above-all approach to messaging about its risks. “There’s a really extreme and egregious lack of accountability and oversight [...] that is going to lead to all sorts of issues,” Krueger told DFD.
AI boosters may face a similar strategic question going into the midterms as they push for fewer regulations so developers can improve AI systems more quickly. Nathan Leamer, who directs the political advocacy group Build American AI, says he’s uninterested in pitching voters on the futuristic utopias that some claim the technology could usher in. Rather, he thinks messaging will be more effective if it emphasizes how AI can revive industrialization and manufacturing in America, or enable tools that help consumers.
“I’m more interested in the tangible benefits right here and now, so I’m not pontificating about what life is going to be like 25, 50 years from now,” he told DFD. “I don’t want to be talking about The Jetsons.”
