Barack Obama has a private warning for Democrats: They badly need to grab hold of the debate over artificial intelligence, and quick. “I would strongly urge that the Democrats put together a framework for a very public conversation,” Obama told a roomful of party donors and lawmakers last week, suggesting that a Democratic-controlled House should be ready to act next year. Given that AI is “moving very fast in private hands,” he warned, inaction now could be “dangerous.”
Obama’s admonition—along with growing fears among AI scientists that unchecked AI development poses existential threats to humanity—seems to have jolted Democrats, unleashing fresh resolve to find the right AI message for the midterms. But I think Obama’s intervention raises a deeper question: What should Democrats really stand for on all this?
I can pinpoint at least three ways that Democrats—with some exceptions—aren’t, as of now, meeting this moment. They aren’t sufficiently harnessing public anger over AI data centers to argue for a clean-energy future. They aren’t stamping themselves as the only party that wants AI’s gains broadly shared in a truly egalitarian way. And they aren’t talking enough about Donald Trump’s role in all of it.
Let’s take the last one first. As Paul Krugman writes, Trump is at the center of this whole sordid story. The short version: as president, Joe Biden signed an executive order calling for a society-wide effort by government, the private sector and civil society to both mitigate the risks and harness AI’s great benefits. But Trump ripped it up and installed a breakneck deregulatory posture, because it’s supposedly necessary to compete with China.
What’s more, Trump himself has flatly dismissed concerns over AI’s extinction-level threat, which he reiterated in crazed tirades on Monday. As Krugman notes, Trump’s coziness with tech titans and his hatred of corporate regulation have produced a deeply reckless stance: “Damn the social and existential risks, full speed ahead.”
There’s another overlooked aspect to all this, however. It’s that Trump is currently using federal agencies to facilitate the development of AI data centers in ways that already harm the public. His Environmental Protection Agency is moving to change federal regulations to make it harder for local communities to get wind of planned data centers before construction starts to frustrate counter-organizing. Trump’s EPA is also weakening other regulations over the data-center permitting process and the gas plants that power them. His Justice Department is actively siding with data-center developers—including Elon Musk—against community lawsuits.
Here again, Trump has snidely dismissed public concerns, insisting that communities who resist data centers will end up “backwards and poor.” Democrats could highlight all this agency activity, but they mostly aren’t. “Democrats can win votes in every corner of the country if they promise their majority will block Trump’s use of the DOJ and EPA as a goon squad to crush local opposition to data centers,” Jesse Lee, senior adviser to the advocacy group Climate Power, tells me.
This doesn’t mean Democrats must absolutely oppose any more data centers no matter what. Some Democrats can back moratoriums; others might not want to. It just means that Democrats can tell voters help is coming: If they win Congress, they will try to stop Trump from using federal agencies to grease the way for tech companies to steamroll them.
True, some Democrats have been running ads against data centers. But most of them—see here and here—treat this as largely a NIMBY issue, stressing higher electricity rates and corrupt local government. A memo circulating among Democratic operatives advises House candidates to tie data centers to corruption and high energy costs, with no mention of Trump’s role in any of it.
I get why Democrats are stressing the NIMBY angle: It’s concrete, tangible, and clear. As tech journalist Jasmine Sun told Substacker Derek Thompson about her reporting on the data-center rebellion, many people do understand this issue in such narrow terms. But there’s an opportunity here to enlarge the scope of the discussion, by connecting data centers to Trumpian oligarchy—both of which are deeply unpopular—in one broader message. As the group Guardrails Action has advised Democrats, public opinion on AI is unusually fluid. So it might be time to attempt bigger arguments.
After all, with data centers already commanding intense public attention, Trump’s smoothing of the process for them opens a window on his effort to facilitate runaway AI development for his corporate cronies in a much broader sense. These are one and the same story. Putting this Trump-oligarch alliance at the center of that story—and linking it to what people see and feel in their own communities—seems like a winner.
Yes, some candidates might fear invoking Trump could trigger negative partisanship among red-leaning voters who are gettable on this issue. But Trump is underwater everywhere, so invoking him is worth considering even in swingy areas. Aaron Rupar flagged this moment from Senator Jon Ossoff—who’s seeking reelection in very-swingy Georgia—in which he did this to good effect:
Ossoff: "On day one, Donald Trump rescinded the AI safety order meant to guard against catastrophic risks, including the engineering of bioweapons. Meanwhile, Don Jr's 1789 Capital bought big money stakes in data centers and AI labs ... President Trump is compromised" [pic.twitter.com/ARTCSRI0ea](https://t.co/ARTCSRI0ea)
[September 13, 2026](https://x.com/atrupar/status/2099232402274967961?ref_src=twsrc%5Etfw)
That’s potent stuff. It captures an observation from Brian Beutler: If there’s good reason to fear AI, it’s in part because of the “devil’s bargain” many tech titans made with Trump. They struck it when his authoritarian power seemed unchallengeable. This itself shows they can’t be trusted with our future. Neither they nor Trump can be. Democrats should tell voters this, because…it’s the plain truth, and they need to know it.
Which brings us to the other themes Democrats should hit. Data-center mania has made voters unusually sensitive to elite self-enrichment via manipulation of public processes. So why not bring in clean energy where appropriate?
For some Democrats, this election won’t be the place to do this. But voters seem unusually tuned into the enormous volume of energy that will be needed to fuel any AI future, and into how oligarchic corruption is streamrolling local democratic aspirations. This comes, as Bill McKibben often notes, even as the Iran war and extreme weather horrors are heightening public awareness of our overreliance on fossil fuels, how it enriches petro-oligarchs, and the continued cheapening of alternative energy. Soon enough it will be time to seize the moment to talk about how to move toward powering the AI future we actually want with sun and wind. Which leads, finally, to the distributional component. In the face of rising fears of AI-driven mass joblessness and exacerbated inequality, Matt Bruenig points out at The Argument that we already have policy tools at the ready to ensure a broadly equitable distribution amid the AI boom. These include socializing some tech profits, doling out dividends to the public from a government-owned investment fund, higher taxes on corporations, expanding the welfare state, and so forth. A more stable, more egalitarian transition is within our reach if we want it.
Perhaps some Democrats disagree with this or that specific idea. But just about every Democrat wants to ensure that the gains of this transition—including whatever major advances it inspires in medicine and other fields—are very broadly shared. Republicans are scrambling to get in sync with public anger by backing some regulation of data centers, and maybe some Republicans sincerely want to act against an AI apocalypse. But there’s zero reason to believe the GOP overall, as a party, has above-board long-term distributional intentions here. Trump has made gestures in the direction of a fairer distribution of tech profits, but he’s also a corrupt liar who’s overseen a very large upward redistribution of wealth, so believe that when you see it.
Now ask yourself: How much current Democratic Party messaging will leave the public aware that it’s far more serious than the opposition about achieving a broadly egalitarian distribution of the fruits of what will be the biggest technological transformation of our lifetimes—and much more devoted to securing truly shared mass improvements in human welfare from it?
In a subsequent statement about his warning, Obama called for a “democratic conversation” about AI’s future, adding: “Whether this technology results in amazing breakthroughs in medicine, energy and education or unleashes huge economic disruptions, greater inequality, and potential catastrophe will depend on the choices that we make right now.”
Obama is calling on Democrats to get on the right side of all that. His admonition isn’t just about the dangers of AI. It’s also a warning against an unnerving possibility—that an addiction to political smallball could muddy who stands where on one of the most consequential moral challenges of our time. It’s an exhortation to go bigger—to tell a much more ambitious story about the future we all want, and about how giving Democrats power will help to get us there.