Who Owns the AI Boom? Public Lands, Private Fortunes and the Conflict at the Heart of America’s Data Center Expansion

The Trump administration’s push to put artificial intelligence infrastructure on federal lands raises troubling questions about financial conflicts, environmental accountability and the public interest.

The race to dominate artificial intelligence is increasingly being presented as a matter of national security. The United States must build more computing capacity, generate more electricity and move faster than geopolitical competitors, the argument goes. But in the rush to achieve technological supremacy, an important question is getting lost: Who gets to decide how America’s public lands and natural resources are used, and who ultimately benefits?

A November 6 investigation by ProPublica‘s Mark Olalde brings this question into sharp focus. It examines Interior Secretary Doug Burgum’s substantial financial interests in artificial intelligence companies while he leads the administration’s efforts to encourage data center development on federal lands.

The investigation is significant not simply because of the potential conflicts of interest it identifies, but because it exposes a much larger problem with the way the AI infrastructure boom is being pursued. Public resources are being positioned to support an industry dominated by some of the world’s wealthiest corporations, while the financial, environmental and social consequences remain inadequately understood.

This is not an argument against artificial intelligence. It is an argument that technological progress should not come at the expense of public accountability.

The financial interests behind the policy

According to ProPublica, Burgum’s 2026 financial disclosure identifies substantial holdings through Arthur Ventures, the investment firm he co-founded. An analysis of the firm’s portfolio identified between $18.1 million and $64.4 million in investments in AI-related companies.

The investigation also reports that Burgum realized between $2.1 million and $11 million in capital gains from AI-related investments during his final year as governor of North Dakota. In 2025, after joining the Trump administration, he reported capital gains of between $2.5 million and $15.1 million from sales of AI-related stocks and equity.

These figures are ranges because federal financial disclosure requirements allow officials to report asset values within specified brackets.

The important distinction is that Burgum agreed to resign his positions at Arthur Ventures but retained his financial interests. The Interior Department maintains that he has complied with federal ethics requirements and denies that his investment holdings influence his official decisions.

There is no evidence in the investigation establishing that Burgum has improperly exercised his authority to benefit his investments. Nor does the existence of an investment automatically establish a violation of federal ethics law.

Nevertheless, the circumstances raise a legitimate question about the appearance of a conflict of interest.

As Cynthia Brown of Citizens for Responsibility and Ethics in Washington explains in the ProPublica investigation, federal ethics principles extend beyond criminal conduct and address situations that could create an appearance of compromised judgment.

Richard Painter, former chief ethics lawyer for President George W. Bush, similarly argues that Burgum should avoid decisions affecting an industry in which he has substantial financial interests.

This is an important distinction. Public confidence in government depends not only on whether officials break the law, but also on whether citizens can reasonably trust that public decisions are being made impartially.

The concern is not confined to Burgum. ProPublica’s examination of financial disclosures found that 380 Trump administration political appointees or their spouses had financial connections to Nvidia, more than 320 had ties to Alphabet, and more than 240 had financial connections to Meta.

These figures do not establish that those officials have acted improperly. They do, however, illustrate how deeply the financial interests of government personnel can overlap with the industries whose development they are helping to accelerate.

Public lands become the next frontier

The geographic scale of the opportunity is extraordinary.

According to a 2026 U.S. Geological Survey report cited by ProPublica, at least 91.3 million acres of federally managed public lands could potentially accommodate data centers.

That is an area comparable in size to Montana.

The Interior Department oversees an enormous portfolio of public assets, including national parks, wildlife habitat, mineral resources, federal lands, dams and responsibilities involving tribal affairs. The Bureau of Land Management alone administers hundreds of millions of acres.

The administration is now considering how some of these lands might be used to accommodate the infrastructure demands of AI.

ProPublica reports that the Bureau of Land Management is reviewing approximately 15 data center applications and 17 related infrastructure projects across Arizona, Nevada, Oregon, Utah, Wyoming and Idaho. Associated projects include power plants and transmission infrastructure.

This development deserves considerably more public scrutiny.

Federal lands are not simply vacant real estate awaiting the highest bidder. They represent an extraordinary public inheritance, supporting recreation, wildlife, watersheds, cultural resources, grazing, mineral development and ecological functions that cannot easily be replaced.

Their management requires balancing competing interests over long time horizons.

A data center, by contrast, is fundamentally a commercial facility designed to generate returns for its owners and customers. Its infrastructure may remain in place for decades, while its technology and economic usefulness can change much more quickly.

That difference should inform the standards applied to federal land leases and development approvals.

The enormous energy and water equation

The environmental implications are not theoretical.

The Lawrence Berkeley National Laboratory’s 2024 United States Data Center Energy Usage Report estimated that American data centers consumed approximately 176 terawatt-hours of electricity in 2023, equivalent to about 4.4% of total U.S. electricity consumption.

Its projections at the time suggested that data centers could account for between 6.7% and 12% of national electricity consumption by 2028.

A subsequent 2025 update, released in June 2026, projects that data centers could consume approximately 11.8% of U.S. electricity by 2030, with a range of 9.5% to 15.3%.

These estimates demonstrate the magnitude of the challenge.

For perspective, electricity demand from data centers is no longer a marginal issue for energy planners. It is becoming a major consideration in decisions about power generation, transmission capacity, grid reliability and household electricity costs.

Water is another critical consideration, particularly in the American West.

A January 2026 fact sheet from the Environmental Law Institute, drawing on Lawrence Berkeley National Laboratory research, estimates that direct water consumption by U.S. data centers increased from approximately 21.2 billion liters in 2014 to 66 billion liters in 2023.

The same research estimated that electricity generation to supply data centers consumed an additional 800 billion liters of water in 2023.

The distinction between direct and indirect consumption matters. A facility may use relatively little water on its own property while relying on electricity generated by power plants that consume substantial quantities of water elsewhere.

The Environmental Law Institute also highlights the particular risks associated with evaporative cooling. Approximately 80% of the water drawn by this cooling method can be lost through evaporation.

These figures have special relevance to proposed projects in the Southwest, where drought, groundwater depletion and competing demands from agriculture, municipalities and ecosystems already create significant pressures.

Technology companies can argue, correctly, that cooling systems are improving. Closed-loop liquid cooling, immersion cooling and other innovations may substantially reduce direct water requirements.

But improved technology does not eliminate the need for transparent, independently verified assessments of total resource consumption.

Nor does it answer the fundamental question of whether a particular location is appropriate for an energy-intensive industrial facility.

The public is not necessarily buying the AI expansion argument

Another important dimension of this debate is public opinion.

The assumption that communities will automatically welcome data centers because they promise investment, jobs and technological progress is increasingly difficult to sustain.

A July 2026 Fox News national poll found that 70% of voters opposed building AI data centers in their communities, compared with 30% who supported them.

A separate August 2026 survey by Public First and the Information Technology and Innovation Foundation found that 46% opposed new data centers in their communities, while just 26% supported them.

These polls used different questions and methodologies, but both point toward substantial public skepticism.

The opposition is understandable. Communities want to know how much water and electricity a facility will consume, whether electricity costs will increase, how many permanent jobs will actually be created, and whether local taxpayers will subsidize infrastructure primarily benefiting private corporations.

There is also a question of democratic accountability.

When a company proposes a large industrial development through local government, residents may have opportunities to participate in zoning hearings, environmental reviews and public permitting processes.

Federal land development can involve a different institutional structure. As ProPublica observes, the public may have fewer opportunities to influence decisions made on federally managed property than decisions made through local and state processes.

That is particularly concerning when projects are being promoted as urgent national priorities.

National security and economic competitiveness are important considerations, but neither should become a blanket justification for reducing transparency.

Are taxpayers receiving a fair return?

One of the most important issues raised by the investigation concerns the financial return to the public.

Companies developing commercial facilities on federal lands generally pay for the right to use those lands. The government receives royalties, rental payments or other fees depending on the nature of the development.

For data centers, the Interior Department has indicated that annual fair-market-value rent will be the principal payment mechanism.

That approach may be reasonable in principle, but it raises a practical question: How will fair market value be calculated for a new and rapidly expanding industry whose infrastructure requirements and commercial value are changing so quickly?

A data center is not necessarily comparable to an ordinary commercial property lease.

It may require extensive transmission infrastructure, substations, water systems, access roads and dedicated power generation. The public may also bear indirect costs associated with environmental degradation, water consumption and the expansion of electrical infrastructure.

The government must therefore look beyond the nominal rental payment.

A proper economic assessment should consider:

• The market value of the land and the duration of the lease.
• The cost of associated infrastructure and who pays for it.
• The project’s direct and indirect water consumption.
• Electricity costs and potential effects on other ratepayers.
• Environmental impacts and long-term restoration obligations.
• Permanent employment, tax revenue and other measurable local benefits.
• The opportunity cost of using public land for industrial development rather than conservation, recreation or other public purposes.

Without this information, it is impossible to determine whether the public is receiving an appropriate return.

The danger is that federal lands become an inexpensive resource for private corporations while the public absorbs a disproportionate share of the costs.

A test case for government ethics

The ProPublica investigation also describes potential financial conflicts involving other senior officials.

Steve Pearce, director of the Bureau of Land Management, has held investments in several major technology companies and is also invested in Titan Development, a real estate company involved in land acquisitions for data centers in the Southwest.

Michael Boren, who has held senior positions at the Interior and Agriculture departments, disclosed financial interests in AI-related companies, including between approximately $1.5 million and $6.1 million in xAI and up to $500,000 in Valar Atomics.

The latter is particularly noteworthy because Valar Atomics is pursuing Project Beehive, a proposed data center complex on thousands of acres of federal land in Utah.

These disclosures do not, by themselves, demonstrate improper conduct. The officials and their agencies have stated that applicable ethics requirements are being followed.

But they reinforce the need for a stronger and more transparent system of conflict management.

At a minimum, officials with substantial financial exposure to companies or industries affected by their decisions should face meaningful independent review, clearly documented recusals where appropriate, and public disclosure of how potential conflicts are managed.

The issue is especially important when the same administration is simultaneously promoting an industry, facilitating its access to public resources and overseeing the regulatory processes governing its development.

What responsible development should look like

There is a legitimate national interest in developing AI infrastructure.

Artificial intelligence has applications in scientific research, medicine, manufacturing, logistics, national security and countless other fields. The United States has good reasons to maintain a competitive position in these technologies.

But there is a difference between supporting technological innovation and providing preferential access to public resources.

A responsible federal policy should begin with several principles.

First, transparency must come before land commitments. All proposed data center leases on federal lands should be accompanied by public disclosure of ownership, financing, expected resource consumption, projected employment and anticipated public revenue.

Second, environmental review must be comprehensive. Water, electricity, transmission infrastructure, emissions, wildlife impacts and cumulative effects should be evaluated before projects receive approval. Assessments should not be fragmented in ways that obscure the combined effects of a data center and its supporting infrastructure.

Third, public participation must be meaningful. Local communities, tribal governments and other affected stakeholders should have genuine opportunities to review proposals and influence decisions before development commitments become difficult to reverse.

Fourth, taxpayers must receive a fair return. Federal leases should be competitively structured where appropriate, periodically reassessed and designed to capture the economic value of public resources.

Finally, conflicts of interest must be addressed proactively. Financial disclosure alone is not sufficient. Where officials have substantial financial exposure to an industry they oversee, independent ethics review and appropriate recusals are essential.

These safeguards should apply regardless of which political party controls the White House.

The larger question: Whose future is being built?

The most important contribution of the ProPublica investigation is that it moves the discussion beyond the familiar debate over whether AI is good or bad.

The real issue is how the infrastructure supporting this technology is financed, located, regulated and governed.

America’s public lands belong to the public. They are not an administrative convenience to be deployed whenever a politically influential industry requires additional space or resources.

The AI industry has the capital, expertise and commercial incentives to build the infrastructure it needs. Government has an important role in establishing predictable rules, supporting research, facilitating responsible energy development and protecting national security.

But government should not become an uncritical development partner for private interests, particularly when senior officials have financial exposure to the very industry they are promoting.

The administration’s AI ambitions may ultimately produce enormous economic and strategic benefits. That possibility deserves serious consideration. So do the costs, the alternatives and the interests of the communities that will live with the consequences.

Technological leadership should not be measured solely by the number of data centers constructed, the amount of computing capacity installed or the market value of the companies involved.

It should also be measured by whether the benefits are broadly shared, whether public resources are protected, whether communities have a voice and whether government decisions can withstand independent scrutiny.

The United States does not need to choose between leading the AI revolution and protecting its public lands.

It needs to insist on both.

The public should not have to surrender its natural resources, its right to participate in decisions, or its confidence in government to finance the next generation of private technological wealth.

---

References and source URLs

Below is a consolidated bibliography with direct links to the original reporting, government research, independent studies and polling data.

1. Original investigative reporting

1. ProPublica — Trump Official Who Wants Data Centers on Public Lands Has Tens of Millions Invested in AI Industry Mark Olalde, November 6, 2026. https://www.propublica.org/article/burgum-public-land-data-centers-investments 

2. Federal government and scientific research

2. U.S. Geological Survey — Colocating Artificial Intelligence Data Centers with Energy Infrastructure on Federal Public Lands Scientific Investigations Report 2026–5035, July 22, 2026. https://www.usgs.gov/publications/colocating-artificial-intelligence-data-centers-energy-infrastructure-federal-public 

3. Full USGS scientific report (PDF) https://pubs.usgs.gov/sir/2026/5035/sir20265035.pdf 

4. Lawrence Berkeley National Laboratory — United States Data Center Energy Usage Report: 2025 Update Published June 2026. https://bies.lbl.gov/publications/united-states-data-center-energy-2025 

5. Berkeley Lab — Report download and research record https://escholarship.org/uc/item/33m6w3x0 

3. Water use and environmental impacts

6. Environmental Law Institute, Tulane Institute for Water Resources Law & Policy — Data Centers and Water Fact Sheet January 2026. Download the original PDF 

7. UC Berkeley, Center for Law, Energy & the Environment — Regulating Data Center Water Use in California February 2026. https://clee.berkeley.edu/publications/regulating-data-center-water-use-in-california/ 

8. Pacific Institute — How Much Water Do Data Centers Use? And Why Does It Matter? September 1, 2026. https://pacinst.org/publication/how-much-water-do-data-centers-use-and-why-does-it-matter/ 

9. Congressional Research Service — Data Centers and Water: Frequently Asked Questions July 31, 2026. https://www.everycrsreport.com/reports/R49057.html 

4. Public opinion and community acceptance

10. Fox News Poll — Cool Your Servers! Voters Reject Data Centers July 27, 2026. https://www.foxnews.com/politics/fox-news-poll-cool-your-servers-voters-reject-data-centers.amp 

11. Public First / Information Technology and Innovation Foundation — Data Centers Are Americans’ Least Favored Form of Development August 26, 2026. https://itif.org/publications/2026/08/26/data-centers-are-americans-least-favored-form-of-development/