Kevin O'Leary cuts Utah data center by 19,430 acres after resident pressure

Kevin O'Leary has agreed to cut his planned Utah data center in half, trimming 19,430 acres from a project that was originally set to cover 40,000 acres. The Shark Tank star made the change in a letter sent Thursday to Utah Senate President J. Stuart Adams, according to reporting by local affiliate ABC4, and the project sits in and around the Locomotive Springs Waterfowl Management Area. For anyone tracking the AI infrastructure boom, that makes this more than a local land-use squabble. It is a very public reminder that giant data center ambitions do not exist in a vacuum. They have to survive politics, water concerns, and the reality that the people living nearby get a vote too.

The timing matters. Adams had just called on O'Leary to slash the size of Project Stratos by 75 percent, which would have brought it down to about 10,000 acres. He also asked O'Leary to implement technology that minimizes water consum… The source cuts off there, but the direction is clear enough: this was not a voluntary light touch from a founder politely refining a site plan. It was a response to mounting pressure. The project had already drawn resistance from residents and activists, and now a top state lawmaker had effectively raised the stakes. Once the political temperature gets that hot, even a celebrity-backed project can find itself negotiating from a much smaller footprint.

That is the bigger lesson for executives watching the data center race. Scale is often the pitch, but scale is also the problem. Large data centers need land, power, cooling, and in many cases significant water resources. Those needs become especially sensitive when a project lands near a protected or environmentally important area, as Project Stratos does in and around the Locomotive Springs Waterfowl Management Area. The source does not spell out every technical detail of the project, and it should not be overread into them. But it does show the basic pressure points that repeatedly show up in infrastructure fights: who controls the land, who pays the environmental cost, and who gets to decide whether the project is worth it.

There is also a political dynamic here that executives should not miss. O'Leary is not just any developer trying to site a facility quietly. He is a high-profile business personality, and Project Stratos was large enough to attract attention well before it had the chance to settle into the regulatory background noise that many industrial projects rely on. Once local residents and activists focus on a project of this scale, elected officials can feel forced to respond. Adams did exactly that by urging a dramatic 75 percent reduction and asking for technology that reduces water consumption. That kind of public intervention changes the negotiation. It signals that a project may need to prove not just economic upside, but social acceptability.

For boards and leadership teams, the practical takeaway is simple: the initial footprint is not the final footprint if local politics turn hostile. The source shows how fast a project can move from ambitious to contested to constrained. O'Leary's response, reducing the site by 19,430 acres, suggests a recalibration toward what is politically and environmentally tolerable. That does not mean the project is dead. It means the economics, permitting path, and community relations now have to work inside a narrower box. And when the box gets smaller, everything inside it gets harder. Land planning tightens. Cooling strategy matters more. Stakeholder management becomes part of the build, not an afterthought.

For other operators chasing AI and cloud demand, the headline lesson is not just that Utah pushed back. It is that the old playbook of announcing a massive build and sorting out the consequences later is getting riskier. Water use is no longer a line item buried in an engineering memo. It is a political issue. Land use is not just a zoning formality. It is a community fight. And when a project becomes symbolic enough, lawmakers can force a re-cut before the first major concrete pour. That is the kind of signal every executive team should notice. Even if the demand for compute keeps climbing, the path to building it is now more exposed, more local, and far less forgiving than many ambitious plans assume.