What if the problem with data centers was never the data centers?
I review tools for a living. Most of my week is spent installing something, breaking it, and then writing down whether it did what the landing page promised. And the single most reliable predictor of whether a tool will disappoint me isn’t the architecture, the pricing, or the benchmark chart. It’s how much the vendor hides before I sign up.
Which is why the current fight over AI data centers reads so familiar to me. According to a Gallup poll, somewhere between 70% and 75% of Americans oppose building large data centers near their homes, and more than half support a national moratorium. Those are not soft numbers. That’s the kind of margin you see when people have already made up their minds, and it suggests the industry has a trust deficit rather than an explanation deficit.
The NDA is the tell
Here’s what’s being floated as a fix: drop the nondisclosure agreements that often accompany negotiations between developers and local governments. On its face, it sounds minor. Procedural. The kind of concession a comms team offers when it wants to look responsive without changing anything material.
I’d argue it’s bigger than that, but not for the reason the industry hopes.
An NDA in a municipal negotiation does something specific: it tells the public that their elected representatives know things they are not allowed to repeat. Water draw. Power commitments. Tax arrangements. Noise. Whatever it is, the residents learn about it after the ink dries. In tool-review terms, that’s a vendor who won’t let you test the product until after you’ve committed to the annual contract. I’ve never once encountered that setup and later found a pleasant surprise waiting inside.
So yes, removing NDAs would help. It removes the structural reason for suspicion. But removing a reason for suspicion is not the same thing as earning trust, and this is where I think the industry is about to misjudge the room.
Transparency isn’t a press release
The pattern I see in software goes like this. A company gets caught being opaque. It announces a transparency initiative. The initiative consists of a blog post, a dashboard nobody asked for, and a page of selectively chosen metrics. Nothing that would actually embarrass anyone makes it through. Six months later the trust numbers are unchanged, and the company concludes that its critics were simply unreasonable.
Dropping NDAs could follow exactly that arc. You can end the NDA and still release a 400-page filing three days before a zoning vote. You can be technically open and practically unreadable. Communities have gotten good at spotting the difference, because they’ve been on the receiving end of it from other industries for decades.
What would actually move the needle looks less like a policy announcement and more like a product change:
- Publishing the terms before the vote, not after, with enough runway for residents to read them
- Stating power and water commitments in plain numbers that a non-engineer can check
- Letting the inconvenient figures out alongside the flattering ones
- Accepting that some communities will still say no, and treating that as a legitimate outcome rather than a messaging failure
That last one is the hard part. Most transparency efforts in tech are designed to produce consent. Real openness has to tolerate rejection, otherwise it’s just a nicer-looking sales funnel.
The politics have already moved
The industry’s window for voluntary fixes is narrowing. In Pennsylvania, Governor Josh Shapiro signed an executive order imposing strict standards on data center development. That’s the shape of what comes next: when private negotiation loses legitimacy, public rulemaking fills the gap. And rules written during a backlash are rarely gentle.
Michael Spencer has described the backlash against AI data centers as accelerating through 2026, spreading rather than settling. The sources I’ve read don’t claim to know whether ending NDAs would be enough to turn that around. I don’t think anyone does. What I’d say with more confidence is that keeping NDAs guarantees it won’t.
What I’d actually watch
If I were scoring this the way I score a tool, I’d ignore every announcement and watch one thing: what happens in the first community that gets full disclosure and votes no anyway. Does the developer walk away and go somewhere else? Or does the openness quietly get rolled back in the next three negotiations because it produced the wrong result?
That’s the test. Transparency that only survives when it wins isn’t transparency, it’s a tactic. And the 70% of Americans who don’t want this in their neighborhood are perfectly capable of telling the two apart.
My honest read? Ending NDAs is the right move and it’s nowhere near sufficient. It’s the minimum entry fee for a conversation the industry spent several years avoiding. Starting the conversation late and starting it under conditions you no longer control are not the same as starting it in good faith, but one of them is still better than nothing.
🕒 Published: