The 2026 midterm elections offered not only a referendum on Trumpism, but a stunning rebuke of the administration’s careless approach to the economy and tariffs, not to mention the President’s picking a fight with Iran that sent energy costs soaring. There is no gray area here, no indecipherable mystery over what voters really wanted from their elected officials. It was a complete and total rejection of the path the country was headed down.
But it was also a condemnation of data centers and, by proxy, Silicon Valley and the AI industry. With 41 data center-related ballot measures in the 2026 midterms across 10 states, nearly every initiative resulted in a ban or moratorium, except for an initiative in Harwood, North Dakota, where voters rejected an independent review “before city approval of industrial developments.”
Folks will be sifting through the meaning of the 2026 midterms for weeks to come, but if one thing was resoundingly clear, it’s that Americans are fed up with the overwhelming insistence that AI, as well as AI compute and data centers, are the future and that you can do nothing to stop it. And while it would be foolish to bet against AI and whatever impact it will have on tomorrow and how it impacts our lives and how we work, it was abundantly clear that the backlash to the industry, but also the tech oligarchy that pushed it in the first place, in its current form, is very much real.
You don’t have to accept some bad things happening for the benefit of AI, and, truthfully, what a remarkably dunderheaded, oblivious, and out-of-touch notion to say out loud.
In other words, voters don’t want this. The story of AI is by no means over, but it became clear on November 4th that we need to start having better conversations about the technology and not one that centers on a message of “adapt or die.” Ultimately, this isn’t a red-or-blue state issue. It’s just the way things are. In an NBC exit poll, 76% of voters said they would oppose building a data center in their community, while 56% said it was a crucial or major factor in the congressperson they voted for. Despite the intense, bipartisan opposition to data centers, President Trump put his full support behind them, saying that communities who opposed them would “end up being backwards and poor” and that we would lose ground to China in the AI race.
Tech executives poured hundreds of millions of dollars into the midterms, and in the end, it only confirmed that people don’t want to live with the loud, incessant hum of a 100-megawatt data center.
Business, as it turns out, has no place in government. It’s one thing to help pave the way for innovation and investing in the country’s future with tax incentives predicated on creating good-paying US jobs; it’s an entirely different thing to hand the keys over to business leaders and give them carte blanche without a skosh of regulatory oversight (and for a problematic industry). Because the people knew they were getting fucked.
Every tech CEO got what they had coming, and we shouldn’t let them forget it either.
Readers of this newsletter are, by now, aware that AI is something I talk about quite a bit, perhaps even to the point where I myself have wanted to let up on the gas at times. However, I would argue that it is necessary to talk about these issues when so much of the federal government has given the AI industry a free pass. Many people have genuine concerns about these tools and technologies, and it’s obvious to anyone with eyes to see that this is a bipartisan issue, with most folks being told to hold their noses and accept their fate.
AI certainly isn’t going anywhere, but what needs to change is not just how these tools are marketed and sold to us for widespread adoption, but the entire industry’s approach to standards and ethics, something that speaks to our concerns around safety, protecting workers, and the wholesale “fair use” theft of IP. The environmental and water-consumption issues alone warrant much-needed regulatory oversight, and most tech companies have failed to be transparent about the water and power required to run these tools.
I wouldn’t necessarily blame you if you’re sick of hearing about AI—I’m just as exhausted as my beloved em dashes that were cruelly taken away from me, and I would love nothing more than to stop writing about it. But when you have a disruptive technology that can potentially upend the entire US economy, and no one in control is raising alarm bells, then we have a very serious problem being handled by deeply unserious people. Frankly, I would much rather spend my time talking about new materials and plastic-free initiatives, or whatever silly watercooler CPG convo is happening around whatever awful flavor Kraft released this week. It’s also my general sense that most rank-and-file designers didn’t go to college to push pixels around on a client’s AI-generated concept for not a whole lot of money, and that the only people who are really talking about how you need to incorporate these tools (at least in this corner of the design world) into your processes are often design leaders, agency heads, and C-suite brand types.
I don’t think most people are against using AI to name files or play a supporting, agentic role, but I’m also sure most of them are sick of having it stare them down on the reg and are rightfully concerned about the direction of their own careers.
The message to every politician, regardless of party, cannot be clearer. Regulation needs to happen now, and the entirety of the AI industry needs to meet real people where they are. We won’t all become millionaires just because we have freed ourselves from the shackles of tedious labor. There will be no 3-day work weeks (though that does have a nice ring to it). Data centers will not be bastions of green energy that don’t wreak havoc on the environment or their community.
And when the 2027 congressional term kicks off in January, hopefully, the watchdogs will finally show up and get to work on what voters actually want.
