How Local Democracy Stopped Wonder Valley
Kevin O’Leary wanted to build the biggest data center in the world in Utah. It’s now stalled, and the people who stopped it weren’t investors or regulators. They were local residents. According to The Verge AI, the project was officially called Stratos but better known as Wonder Valley, a nod to O’Leary’s Shark Tank nickname. It was planned as a 40,000-acre AI campus drawing nine gigawatts of power. That’s more than double the average power use of the whole state of Utah.
Verge features writer Josh Dzieza spent months reporting on the project. He found that the way it was planned, announced and approved set off a political firestorm that “stopped Stratos in its tracks.” The fallout keeps growing. O’Leary called his critics Chinese spies on Fox News, Fox had to apologize on air, and a defamation lawsuit followed.
Here are the key questions behind the story and what they mean for the AI infrastructure boom.
How did a project this big stay hidden?
According to Dzieza, almost nobody knew about it. For months, only a handful of property owners in a remote valley, a few people at one state agency and some real estate developers were in the loop. Most people in the county and the state first heard about it in late April, when O’Leary started talking it up in the media.
What stands out here is the order of events. The public heard the sales pitch before it heard about the plan. When residents feel they’ve been handed a done deal, the anger usually goes after the whole project, not just the details.
Who gave the green light?
The project went through MIDA, the Military Installation Development Authority. It’s a state development agency meant for projects with some military purpose. Dzieza notes that purpose “can be fairly tangential”: MIDA also runs a luxury ski resort that offers discounts to veterans.
MIDA’s real draw for developers is its power. Inside a project area it sets up, it effectively becomes the municipal government and can override local land-use rules. That could speed up the permitting and zoning that slow down so many data center projects. To set up a project area, though, MIDA’s board has to ratify it, every landowner has to consent, and the local government, here the county, has to sign off as well.
That’s the core tension. Agencies built for speed can push projects through, but they also skip the public process that gives a project legitimacy.
Is this just Foxconn all over again?
Partly. Dzieza previously covered Foxconn’s promised LCD factory in Mount Pleasant, Wisconsin. It was supposed to bring 13,000 jobs that never showed up. That land now hosts a Microsoft data center.
He didn’t expect the comparison to fit. Unlike Foxconn’s factory, the data center buildout is real, and many of these projects actually get built. But he found more parallels than he expected: the same mix of local officials, planning meetings and angry residents. He also found more data centers that exist mostly “on paper and speculation” than he’d realized, and it’s often hard to tell which ones are real.
Why does this matter right now?
The AI industry is spending huge sums on compute, and power and land are the bottlenecks. Wonder Valley shows a third one that money can’t fix: local democracy. The backlash against data centers is also bipartisan, which makes it harder to brush off as one side’s politics.
A few broader signals:
- Scale invites scrutiny. A project that uses more power than an entire state will get attention from voters, not just utilities.
- Speculative announcements carry risk. When headline projects fall apart, people grow more skeptical of the ones that are real.
- Fast-track agencies can backfire. Getting around local review may save months, then cost the whole project.
What should builders and buyers do?
If you develop, finance or plan to lease AI capacity, these steps follow from the story:
- Engage the community early. Residents shouldn’t learn about a project from a TV interview.
- Stress-test the approval path. Know which consents you still need, including county sign-off.
- Check how real a project is before committing. Look for actual power agreements, land control and approvals, not just press coverage.
- Watch the messaging. Attacking critics, as O’Leary did, can turn a local fight into a national story and a lawsuit.
The infrastructure race will keep going, but Wonder Valley shows that permission from local communities is now a real constraint. Companies that treat it as a box to check will keep running into trouble. Dzieza’s full reporting on how the project unraveled is worth reading at The Verge.