What is actually on the table
Land records reviewed by the Center for Western Priorities, a non-partisan conservation group, show the Bureau of Land Management is weighing 35 data center-related applications on public land across the western US. Seventeen are proposals for full data center campuses, thirteen cover supporting AI infrastructure such as power generation or networking buildout, and five could not be classified from the available filings. The group published a tracker and map on October 22 built directly from BLM land records, giving outside observers the first consolidated public view of how much federal land is now in play.
The standout example is an eastern Utah project covering more than 9,000 acres, designed to run on small modular nuclear reactors rather than grid power. That acreage figure matters on its own: a typical hyperscale campus runs a few hundred to low thousands of acres at most. A 9,000-acre footprint signals either an extraordinarily ambitious single site or a reservation of land well beyond near-term construction needs, and either reading should raise real questions for anyone evaluating whether that capacity will actually materialize on the timeline a developer promises investors or tenants.
Why public land is the newest frontier
Private and state land near population centers has become expensive and politically contested, as the moratorium fights in Virginia and Memphis over the past year have shown clearly. Federal land in the sparsely populated western US offers an alternative: fewer immediate neighbors to object at a town hall, and in theory a more predictable federal permitting path than a patchwork of county zoning boards each with their own politics. That combination is clearly attractive enough that developers are now filing applications at a volume BLM was never built to process at this scale or speed.
Lauren Bogard, the Center for Western Priorities' deputy director, put the core objection plainly: data centers do not need public land the way mining or drilling operations genuinely do, since they are not extracting a resource that exists only there and nowhere else. Her argument is that siting a data center on federal land is a policy choice, not a resource necessity dictated by geology, and it deserves a dedicated environmental review and public input process rather than being waved through under frameworks originally built for extractive industries like mining and oil drilling.
The industry's counter-numbers
The US Chamber of Commerce released its own report the same week, framing data centers as economic engines rather than a federal land giveaway to private industry. The Chamber's figures claim a single data center's two-year construction phase can support an average of 5,200 jobs and add $680 million to a state economy, generating roughly $400 million in annual state personal income over that period. It cites five-year tax revenue impacts of $96 million to $174 million across counties in Texas, Virginia, Georgia, and Ohio, and estimates a typical 50MW facility adds 750 construction workers and 740 permanent jobs annually once operational.
Both sets of numbers are largely talking past each other rather than engaging directly. The Chamber's figures describe aggregate economic activity generated once a project is built, not who bears the environmental review burden or who gets a meaningful say before a 9,000-acre lease is finalized on land the public technically owns. The conservation group's objection is fundamentally about process and consent, not whether data centers generate real jobs and tax revenue. Enterprises watching this fight from the outside should recognize that neither side's numbers actually resolve the operative question: how fast, and how durably, will BLM-approved capacity get built and stay built.
The Trump dividend context
This tracker surfaced one day after President Trump suggested data center developers pay local communities a direct dividend to blunt opposition to new construction, a proposal aimed squarely at the private and state-land fights dominating headlines through the midterm election season. Federal land applications sit outside that framework entirely, since there is no local tax base or directly affected resident population in the same way a suburban county has one. That gap is exactly where a watchdog group like the Center for Western Priorities sees the sharpest risk: no dividend mechanism exists for public land, no established local review body has jurisdiction, and a federal agency is now processing a volume of industrial land requests it has never handled before at this scale.
Expect this to become a live political issue well before any of the 35 applications reach a final construction decision. Public land management draws scrutiny from a different and often more litigious set of advocacy groups than local zoning fights typically do, and environmental review timelines on federal land can run longer than state-level permitting even when there is no organized local opposition to a specific project at all, simply because the federal review process has more statutory steps built into it.
What this means for capacity planning
If your AI infrastructure roadmap includes capacity that a provider is sourcing from a federal-land site, build in a longer and less predictable timeline than you would for a comparable private-land deal, and ask your provider directly how they are hedging against permitting delay or legal challenge from groups with a documented track record of filing them. A 9,000-acre nuclear-powered Utah campus sounds compelling in a vendor pitch deck, but it is also precisely the kind of project most likely to attract sustained legal scrutiny from a well-resourced conservation group.
The more durable lesson is that data center siting has become a genuine due diligence category for enterprise buyers, not a background detail to take on faith from a provider's marketing materials. Ask where the capacity physically sits, who legally controls the land, what review process it went through before approval, and what contractual protection exists for your committed capacity if that review stalls or gets challenged in court. The answer increasingly separates providers with real delivery discipline from those betting on land they have not yet secured the clear legal right to build on.
