Utah has taken a major step toward hosting what could become one of the largest data centers on Earth, and the decision has landed in the middle of a fight that is no longer just about technology. It’s about water, land, power grids, local governance, and the question of who gets to decide what “progress” looks like when the costs are measured in drought risk and electricity demand rather than headlines.
Earlier this month, Box Elder County commissioners approved the Stratos Project, a planned 40,000-acre data center sprawling across Hansel Valley. The scale alone is difficult to grasp: the project is described as more than twice the size of Manhattan, and it is being positioned as a cornerstone of an effort to expand American AI infrastructure. But as the approval moves forward, so do the concerns—raised by experts, environmental advocates, and residents—about how such a facility would affect an already stressed region, particularly in relation to water availability and environmental impacts.
The Stratos Project is backed by Shark Tank investor and venture capitalist Kevin O’Leary, whose involvement has helped bring national attention to a development that might otherwise have remained a local planning story. Yet the controversy isn’t driven by celebrity investors. It’s driven by the basic physics of data centers: they require enormous amounts of electricity, and they require cooling systems that can translate into significant water use depending on the design. In a place like northern Utah—where climate variability and water constraints are not abstract concepts—the difference between “estimated” and “experienced” impacts matters.
To understand why the approval is so contentious, it helps to look at the three pressure points that keep returning in coverage and public discussion: power, water, and land.
Power: the grid question behind the AI promise
The Stratos Project is projected to consume 9GW of power. That figure is not a minor addition to the energy landscape; it’s a statement about the kind of infrastructure the project intends to build. For context, the reporting around the proposal notes that 9GW is almost double Utah’s peak electricity demand referenced in the coverage. Even if the exact comparison depends on the specific dataset used, the underlying point remains: a facility of this magnitude doesn’t simply “use electricity.” It changes how utilities plan for generation, transmission, and reliability.
Large data centers are often described as flexible loads—machines that can, in theory, be scheduled or throttled. But in practice, the operational requirements of modern AI workloads tend to push toward continuous or near-continuous operation. That means the power demand is not just large; it is persistent. And persistent demand is what stresses grids, especially during heat waves or other periods when electricity demand spikes.
There’s also the question of timing. Data centers can be built quickly compared to the time it takes to expand transmission lines or bring new generation online. When a project is approved before the full energy pathway is secured, communities can end up with a “paper promise” of future capacity while the grid bears the burden of uncertainty in the meantime. That uncertainty is part of what fuels backlash: residents may support economic development, but they don’t want to be asked to accept reliability tradeoffs or higher costs without clear answers.
Water: cooling at industrial scale
If power is the grid issue, water is the regional issue. The Stratos Project has been criticized for potentially straining water supplies in a region where water is already a limiting factor. The reporting highlights concerns about environmental damage and water demands, with references to expert warnings and estimates of consumption.
Data centers need cooling because servers generate heat continuously. Cooling can be achieved through multiple methods—air cooling, evaporative cooling, closed-loop systems, and hybrid approaches—but the key point is that the cooling strategy determines whether water becomes a central constraint. In arid or semi-arid regions, even “efficient” systems can still translate into meaningful water withdrawals or evaporation losses at scale.
This is where the debate becomes more than a technical argument. For many residents, water is not just a resource; it’s a shared foundation for agriculture, ecosystems, and long-term resilience. Northern Utah’s environment is shaped by snowpack, seasonal flows, and the delicate balance between human use and natural systems. When a project of this size enters the conversation, people understandably ask: what happens to the baseline? What happens during drought years? What happens to downstream users?
The Stratos Project’s proponents argue that the facility will be designed to meet regulatory requirements and that the region can accommodate growth. Critics counter that the scale of the project makes “regulatory compliance” insufficient as a reassurance. Compliance can mean meeting minimum standards, but it doesn’t necessarily address cumulative impacts—especially when multiple large developments are competing for the same limited resources.
Land use: the visible footprint versus the hidden footprint
A 40,000-acre site is a visible footprint, but the impacts of a data center extend beyond the land it occupies. Construction requires materials, roads, workforce housing, and ongoing logistics. The facility also requires supporting infrastructure: power lines, substations, water pipelines or wells, and potentially new transmission corridors. Each of these elements can fragment habitats, alter drainage patterns, and change the character of a landscape.
Hansel Valley is not an empty canvas. It’s part of a living region with ecological value and community identity. When a project is described as “more than twice the size of Manhattan,” it’s easy to focus on the number and miss the lived reality: that acreage represents miles of change—grading, construction, and long-term industrial operations.
Critics worry that the project’s scale will overwhelm local planning processes. Even if each individual permit is justified on its own terms, the combined effect can be transformative. Supporters may see the development as a necessary step toward AI leadership and economic growth. Opponents may see it as a gamble with environmental stability.
The unique twist in this story is that the approval came from county commissioners. That means the decision is rooted in local governance, not just state or federal policy. Local officials are often tasked with balancing economic development opportunities against the risks of environmental harm and infrastructure strain. In this case, the commissioners signed off on the Stratos Project despite warnings and backlash, effectively moving the project from concept and negotiation into a more concrete phase.
Why Kevin O’Leary’s involvement matters—and why it doesn’t settle the debate
Kevin O’Leary’s backing has made the Stratos Project harder to ignore. His name signals investment confidence and suggests the project has attracted serious capital. For some observers, that translates into credibility: if a high-profile investor is involved, perhaps the project has done its homework.
But investment backing doesn’t resolve the core questions. Data centers are not only financial ventures; they are physical infrastructure. The feasibility of a 9GW facility depends on power availability, transmission upgrades, and cooling design. The environmental impact depends on water sourcing, discharge or evaporation patterns, and land disturbance. Those factors are not solved by a check—they are solved by engineering, permitting, and long-term monitoring.
In other words, O’Leary’s involvement may explain why the story went national, but it doesn’t answer the concerns that residents and experts are raising. The debate is ultimately about whether the region can absorb the project’s demands without unacceptable tradeoffs.
The “AI dominance” framing and the public’s skepticism
The Stratos Project is described as part of a push to strengthen American AI capabilities. That framing is common in large-scale AI infrastructure proposals: the argument is that the country needs more compute, more data center capacity, and more energy to compete globally.
Yet public skepticism often emerges when the benefits are framed broadly while the costs are localized. Communities may be asked to host infrastructure that supports national competitiveness, but the environmental and resource impacts are borne locally. That mismatch can create political friction even among people who generally support technological progress.
There’s also a deeper concern: whether the promised benefits will materialize in ways that matter to the host community. Data centers can bring jobs during construction and some ongoing employment, but they don’t always deliver the kind of diversified economic development that communities hope for. If the project’s primary value is compute capacity for external customers, residents may feel they are subsidizing national goals without receiving commensurate local gains.
That doesn’t mean data centers never benefit communities. They can support local tax bases, attract related industries, and create specialized employment. But the scale of Stratos—and the intensity of the resource debate—means the public is asking for more than generic promises. People want specifics: what jobs, what timelines, what mitigation measures, what monitoring, and what happens if projections change.
What “approval” actually means next
One reason these stories become so heated is that “approval” can sound like a final verdict, when in reality it’s often a step in a longer process. When commissioners sign off, the project typically moves into further permitting, design refinement, and infrastructure planning. That includes additional reviews that may address environmental impacts, water sourcing, and power delivery.
However, the approval itself is still significant. It signals that local officials are willing to proceed despite warnings. It also shapes expectations for the developer and for future stakeholders. Once a project is on a path forward, it can become harder for opponents to stop it later, even if new information emerges.
This is why the backlash matters now. Public scrutiny early in the process can influence conditions attached to permits, require additional studies, or push for mitigation commitments. If the project proceeds without strong safeguards, the community may find itself reacting to decisions already locked in.
A unique lens: the region’s resilience is the real stake
The most compelling way to view the Stratos Project controversy is not as a simple “pro vs anti data center” divide. It’s a question of resilience. Northern Utah’s water systems and ecosystems are resilient within certain ranges. Large industrial projects can push those systems toward thresholds—especially under drought conditions, extreme heat, or unexpected demand growth.
Even if a project’s average-case estimates look manageable, the risk is in the tails: the years when conditions are worse than expected. Water planning often struggles with uncertainty because the consequences of