The Open Record L3C
March 3, 2026
By Angela Fisher
BOTTOM LINE UP FRONT
On February 27, George Washington University announced it had sold its Virginia Science and Technology Campus to an undisclosed buyer. The 850+ faculty and staff learned their campus was gone. No public hearing. No community input. Just a done deal wrapped in “confidentiality provisions.”
Four days later, the buyer was revealed: Amazon Data Services. The price: $427,299,350.
But here’s what makes this sale remarkable: Amazon still needs a special zoning exception to build data centers on the property. Loudoun County Supervisor Juli Briskman said she’ll “fight hard” against approval. The county has enough data centers, she said. The area is power-constrained. More housing is needed, not more server farms.
Except Amazon already owns the land. The community gets to fight, but from a much weaker position. The sale is complete. The money is paid. The deed is transferred.
Compare this to communities that have successfully organized against data centers throughout 2025. Township boards held public hearings. Residents packed meetings. Legislators introduced bills. Communities won moratoriums, forced ballot initiatives, blocked projects.
So hyperscalers are adapting. Buy universities instead. Skip the public process. Fight zoning after purchase, from a position of ownership.
Two universities are testing this strategy right now:
Virginia (GW/Amazon): Private sale, no public process, $427M paid before seeking zoning approval.
Michigan (U of M/Los Alamos): Public university partnership, $100M state grant, zoning exemption, federal classified research.
Different mechanisms. Same goal: Convert educational assets to data infrastructure while minimizing community input.
If these aren’t test cases, the timing is remarkable. If they are, dozens of universities meet the same vulnerability criteria. And the community resistance mechanisms that worked in 2025 may not work in 2026.
THE VIRGINIA MODEL: Buy First, Ask Permission Later
What Happened
February 27, 2026: GW President Ellen Granberg announced the sale in a letter to the community. The buyer? Confidential. The price? Confidential. The terms? We’ll share details “as they are developed.”
March 2, 2026: The GW Hatchet, the student newspaper, obtained the deed records from Loudoun County. Amazon Data Services. $427,299,350. Roughly $3.5 million per acre for 122 acres.
That evening, GW spokesperson Julia Garbitt confirmed the sale to multiple news outlets.
The community learned about a $427 million transaction affecting 850+ people after the deal closed.
What’s Being Displaced
This isn’t a minor satellite office. GW’s Virginia Science and Technology Campus houses:
The entire School of Nursing. All 800 nursing students attend classes at VSTC. The campus has a 20,000 square foot Simulation Learning and Innovation Center with high-fidelity mannequins and VR immersive learning experiences. It’s ranked as one of the nation’s best nursing programs.
17 research labs, centers, and institutes. The Earthquake Engineering and Structures Laboratory houses one of the largest earthquake simulators on the East Coast. The National Crash Analysis Center studies vehicle safety. Transportation safety research. High-performance computing. Sustainable energy research.
GW’s main data center. The irony is almost perfect. GW is selling its data center to Amazon to build a bigger data center.
850+ faculty and staff. Administrative offices operate from VSTC: the Controller, the Registrar, the VP of Finance and Treasurer. About 250 administrative staff work there daily.
20+ degree and certificate programs. Cybersecurity. Education. Engineering. Health sciences. Information technology. Nursing.
All of this gets five years to relocate. GW will “ideally” move the nursing school to the Foggy Bottom Campus. Plans for the earthquake simulator, the crash analysis center, the research labs? “We’ll figure that out in the coming months,” CFO Bruno Fernandes told the GW Hatchet. “We couldn’t do a lot of planning before making this announcement because of the confidentiality issues.”
The Financial Pressure
GW didn’t have to sell. The university has a $24 million structural deficit. Significant, but not catastrophic. Other options existed:
- Program consolidation
- Tuition increases
- Enrollment growth strategies
- Administrative overhead cuts
But Loudoun County commercial land values rose 55 percent in one year. The average price per acre hit $3.76 million in 2026, a $1.4 million increase since 2025.
CFO Fernandes said officials “wanted to capitalize on” the rising prices. Several buyers approached GW about 18 months ago.
The deficit created vulnerability. The data center boom created opportunity. The sale wasn’t inevitable, it was chosen because it was lucrative and to capitalize on the current market value.
The Process That Wasn’t
No public hearing. No community vote. No township approval. Private universities can sell real estate to whoever they want, whenever they want, for whatever price they negotiate.
The community’s only leverage comes afterward: zoning.
And here’s where it gets interesting. Under a 2023 Loudoun County ordinance, the VSTC property requires a special exception to build data centers. Amazon owns the land, but still needs county approval to convert it.
Loudoun County Supervisor Juli Briskman responded the day the sale was confirmed: “I was disappointed to hear about the sale. I will fight hard not to have more added in that area because the county has enough data centers and should diversify its economy, the area is power-constrained because of the data centers, and more housing is needed.”
The community gets to fight. But Amazon already owns the property. Amazon already paid $427 million. Amazon has five years with GW as a tenant generating rental income. Amazon can afford to wait out opposition, modify plans, negotiate concessions – all while owning the asset.
The leverage shifted before the public even knew a sale was happening.
The Tax Subsidies (Automatic)
Virginia’s Data Center Retail Sales and Use Tax Exemption program cost the state $1.9 billion in the current fiscal year – 81 percent of all state economic incentive spending. That’s for all data centers statewide, not any single project.
The program is automatic. Meet the requirements – $150 million investment and 50 new jobs – and the exemptions kick in. No discretionary approval. No public hearing. Companies qualify “as of right.”
The benefits include:
- Sales and use tax exemption on all equipment and software
- Property tax exemption on servers and equipment
- Extensions available up to 15 years beyond the program’s 2035 sunset
Amazon already operates over 60 data centers in Loudoun County. The company qualifies for the exemption program across those facilities. The GW campus conversion will add to that total.
Based on typical data center equipment costs and the 4.3% Virginia sales tax rate, the GW facility conversion could generate an estimated $100+ million in tax exemptions for Amazon over 15 years. That’s roughly 23% of the $427 million purchase price – paid back through automatic tax breaks.
The exemption program has grown 1,051 percent since creation. It increased from $685 million in FY 2023 to $1.022 billion in FY 2024 to $1.9 billion in the current fiscal year. And it keeps growing as more data centers come online.
Virginia doesn’t disclose which companies receive how much in exemptions. But the scale is clear: $5.2 million per day in foregone revenue, statewide, with no community vote required.
THE MICHIGAN MODEL: Bypass Through Exemption
What’s Happening
December 2024: The Michigan Strategic Fund approved a $100 million grant to the University of Michigan for a partnership with Los Alamos National Laboratory. Total project cost: $1.2 billion. U of M contributes $850 million. Los Alamos contributes $300 million. Michigan taxpayers contribute $100 million.
December 2025: State Representative Jimmie Wilson Jr. introduced HB 5362 to rescind the grant. Seventeen bipartisan co-sponsors signed on. Ypsilanti Township passed a resolution opposing the project. The City of Ypsilanti passed a resolution opposing the project. Over 800 U of M employees, faculty, and students signed a petition demanding cancellation.
March 2026: The grant remains in place. The site is still under review. Construction is planned for 2027 if approved.
The Deception
U of M told state officials the project required 20 acres.
U of M purchased 124.68 acres – more than six times what they disclosed.
Ypsilanti Township officials said the university “misled state officials about the size of the project.” The township wants U of M to locate the facility at the American Center for Mobility instead of the Textile Road site, which sits near the Huron River watershed.
U of M is “reevaluating” the ACM location after initially ruling it out.
The Numbers
Jobs promised: 200 total. Only 30-50 will work on-site.
Power demand: 100-110 megawatts.
Los Alamos National Laboratory’s 2026 budget request: 84 percent for nuclear weapons work.
This is a federal classified research facility. The petition from U of M employees notes that Los Alamos director Thom Mason said in 2019: “Los Alamos is clearly, front and center, a national security lab.”
U of M says the facility won’t manufacture weapons. U of M does not rule out whether research performed there could advance weapons technology.
The Exemption
U of M is a public university. Public universities are exempt from local zoning in Michigan.
U of M doesn’t pay property taxes.
Ypsilanti Township can pass resolutions. The township can organize opposition. The township can demand alternative sites. But the township cannot legally stop the project through zoning authority.
Only the state legislature can rescind the $100 million grant. HB 5362 is in the House Appropriations Committee. Representative Wilson is seeking a hearing in early 2026.
The university bypasses local control entirely. Not through private purchase, but through public exemption.
The Resistance
This is where Michigan differs from Virginia. The public process creates opportunities for opposition:
Legislative action: HB 5362 has 17 bipartisan co-sponsors. The grant can potentially be rescinded.
Public petitions: 800+ U of M employees, faculty, and students signed a petition opposing the project.
Local resolutions: Both Ypsilanti Township and the City of Ypsilanti passed resolutions opposing the Textile Road site.
Transparency: The $100 million grant is public record. The project details are disclosed (even if U of M misrepresented the acreage). The community can see what’s happening and organize accordingly.
Compare this to Virginia, where the community learned about a $427 million sale after it closed.
Michigan’s process is flawed. The university’s zoning exemption means townships lack real power. But the process exists. The grant can be rescinded. The site can be relocated. The project can potentially be stopped.
Virginia’s process doesn’t exist. The sale is complete. The fight is over zoning requirements after Amazon already owns the land.
THE PATTERN: Different Mechanisms, Same Goal
| Aspect | Michigan (U of M) | Virginia (GW) |
|---|---|---|
| Institution type | Public research university | Private research university |
| Partner/Buyer | Los Alamos National Lab (DOE) | Amazon Data Services (AWS) |
| Total cost | $1.2B ($100M state grant) | $427M (direct purchase) |
| Public approval | State grant (public process) | Private sale (no process) |
| Community voice | Resolutions, legislative pushback | Zero (learned after sale) |
| Zoning authority | Exempt (public university) | Must seek special exception |
| Transparency | Grant disclosed (enabled opposition) | Confidential until deed obtained |
| Reversibility | Possible (HB 5362 pending) | Very difficult (sale complete) |
| Current status | Under review, facing resistance | Sale done, zoning fight pending |
Both cases share:
- Educational institution under financial pressure
- Valuable land near data infrastructure
- Public subsidies (grants or automatic tax exemptions)
- Educational mission subordinated to real estate value
- Students, faculty, and communities displaced
The fundamental difference:
- Michigan has a bad public process. But the process exists.
- Virginia has no public process at all.
Both test what communities can do when data center developers bypass traditional approval mechanisms.
THE FEDERAL CONTEXT: Executive Orders and Public Lands
The university bypass strategies aren’t happening in a vacuum. They align with explicit federal policy directives issued in 2025.
Executive Order: “Accelerating Federal Permitting of Data Center Infrastructure”
On July 23, 2025, President Trump issued an executive order directing federal agencies to “facilitate the rapid and efficient buildout” of data center infrastructure. The order establishes a national priority: AI data centers and the power infrastructure to support them.
The key directives:
Federal lands deployment:
- Department of Energy, Department of Interior, and Department of Defense instructed to “authorize data center construction on appropriate Federal lands”
- Military installations: “competitively lease available lands” for projects supporting DOD energy and mission needs
- Brownfield and Superfund sites: EPA to develop guidance for “expeditious” reuse as data center locations
Financial support:
- Secretary of Commerce directed to provide loans, grants, tax incentives, and offtake agreements
- “Qualifying Projects” defined as: $500 million+ capital investment OR 100+ megawatts electric load OR projects protecting national security
Streamlined environmental review:
- Federal agencies to identify NEPA categorical exclusions to “facilitate construction”
- New categorical exclusions to be created for projects “that normally do not have a significant effect on the human environment”
- Federal financial assistance representing less than 50% of project costs “shall be presumed not to constitute a major Federal action” under NEPA
Expedited permitting:
- Projects designated for FAST-41 framework (coordinated federal review)
- EPA to “assist in expediting permitting on Federal and non-Federal lands”
The day after the executive order, July 24, 2025, the Department of Energy announced four federal sites selected for AI data center development: Idaho National Laboratory, Oak Ridge Reservation, Paducah Gaseous Diffusion Plant, and Savannah River Site. Site-specific solicitations expected in coming months.
The Los Alamos Connection
Los Alamos National Laboratory is a Department of Energy facility. The U of M partnership announced in December 2024 – seven months before the executive order formalized the policy.
But the policy direction was already clear. The DOE issued a Request for Information in April 2025 seeking industry feedback on “developing AI infrastructure on DOE-managed lands.” The U of M/Los Alamos deal aligned with federal priorities before they were formally codified.
The project fits the executive order’s definition of a “Qualifying Project”:
- National security focus (Los Alamos is “clearly, front and center, a national security lab”)
- 100+ megawatt electric load (project requires 100-110 MW)
- Federal facility partnership (DOE’s Los Alamos)
The Private University Bypass
The executive order doesn’t just promote federal lands. It explicitly directs EPA to expedite permitting on “Federal and non-Federal lands.”
Amazon’s purchase of GW’s campus doesn’t involve federal land. But it aligns with the broader federal AI infrastructure buildout priority. And Virginia’s automatic data center tax exemptions, $1.9 billion per year, function as the state-level equivalent of the federal financial support the executive order promotes.
The federal government is saying: Build AI infrastructure. Expedite permitting. Use public lands. Provide financial support. Treat this as national priority.
Universities, public and private, are responding. Federal facilities partner with public universities that have zoning exemptions. Private universities sell to hyperscalers that qualify for automatic tax breaks.
Different mechanisms. Same alignment with federal AI infrastructure priorities.
The Timing Pattern
December 2024: U of M announces $100M partnership with Los Alamos National Laboratory.
April 2025: DOE issues Request for Information on AI infrastructure at DOE-managed lands.
July 23, 2025: Trump signs executive order on accelerating data center permitting.
July 24, 2025: DOE announces four federal sites for AI data center partnerships.
February 2026: GW sells Virginia campus to Amazon Data Services for data center development.
The university deals bracket the executive order. U of M/Los Alamos precedes it. GW/Amazon follows it. Both align with the federal directive: Deploy AI infrastructure rapidly using available mechanisms to bypass traditional approval processes.
ARE THESE TEST CASES? THE EVIDENCE
We cannot know for certain whether these represent coordinated strategy. University real estate decisions and federal partnerships happen for many reasons. But examining the pattern suggests this is more than coincidence.
Criterion 1: Timing Synchronization
U of M’s grant: Approved December 2024.
GW’s sale: Announced February 2026.
Both projects are in active phases simultaneously. Both are facing peak resistance in March 2026. These aren’t isolated decisions made independently over years. These are parallel strategies deployed in a narrow window.
Criterion 2: Mechanism Differentiation
Michigan tests: Public university + federal partner + state grant + zoning exemption.
Virginia tests: Private university + tech company + direct purchase + post-sale zoning fight.
These are fundamentally different approaches. If the goal were simply to build data infrastructure, why test two completely different mechanisms at the same time? Strategic testing evaluates alternatives to determine which faces less resistance.
Criterion 3: Resistance Type Testing
Michigan faces legislative opposition. The grant can potentially be rescinded through HB 5362.
Virginia faces zoning opposition. Amazon must seek a special exception after already owning the property.
These test different types of community power. Which is easier to overcome: legislative reversal or zoning approval from a position of ownership?
Criterion 4: Market Coverage
Michigan represents the Midwest, public university systems, state-level approval processes.
Virginia represents the East Coast, private universities, minimal approval requirements.
Testing whether these mechanisms work across different regulatory environments matters if the goal is to scale the strategy nationwide.
Criterion 5: Replicability
Both involve conditions that exist at dozens of other universities:
- Financial pressure creating desperation
- Valuable land in or near data center corridors
- Community resistance to traditional data center proposals
- Need for alternative deployment strategies
If these are test cases, the vulnerabilities they exploit aren’t unique to U of M and GW.
Probability Assessment
If these are coincidences:
- Two universities independently decided to partner with tech/federal entities for data infrastructure within 14 months
- Different approaches (grant vs. sale, public vs. private) happened to emerge simultaneously
- Both reached peak resistance in March 2026 by chance
- Similar patterns at other universities (Brown, UChicago) are unrelated
If these are test cases:
- Coordinated strategy testing different bypass mechanisms
- Deliberately designed to evaluate which approach faces less effective resistance
- Timing allows comparison of outcomes before scaling successful models
- Broader pattern indicates systematic strategy being deployed
Our assessment: The probability these represent coordinated strategy testing is significantly higher than coincidence.
The timing, the mechanism differentiation, and the broader pattern of universities under financial pressure “monetizing real estate holdings” all point to a systemic approach rather than isolated decisions.
WHO’S NEXT: THE VULNERABILITY CRITERIA
If these are test cases, or even if they’re not, the conditions creating vulnerability exist at dozens of institutions.
Identification Criteria
Financial Pressure Indicators:
- Structural deficit (not one-time shortfall)
- Recent budget cuts (FY2026-2027)
- Layoffs or program closures
- Explicit mention of “real estate monetization”
Geographic Indicators:
- Located in or adjacent to data center corridors
- Near existing hyperscaler infrastructure (AWS, Microsoft, Google)
- Land values increasing rapidly
- Power and fiber infrastructure already present
Institutional Vulnerability:
- Public university (can use zoning exemption) OR private university (can sell without approval)
- Satellite campuses or “non-strategic holdings”
- Declining enrollment projections
High Probability Targets
Brown University (Providence, Rhode Island)
Deficit: Undisclosed, but cutting 48 jobs plus 55 vacant positions.
Direct quote from September 2025 announcement: “Plan to monetize non-strategic real estate holdings.”
Institution type: Private (can sell without approval).
Assessment: 90% probability. They literally announced their intention to monetize real estate.
University of Maryland System (Multiple campuses)
Deficit: $155 million state funding cut for FY2026.
Location: Borders Virginia Data Center Alley in Loudoun County.
Institution type: Public (zoning exemption available).
Assessment: 85% probability. Perfect geographic positioning plus severe financial pressure.
University of Chicago (Chicago, Illinois)
Deficit: $160 million structural deficit (down from $288 million).
Location: Chicago has 224 data centers. The highest concentration in the United States.
Institution type: Private.
Assessment: 75% probability. Massive deficit in the largest US data center market.
Temple University (Philadelphia, Pennsylvania)
Deficit: $60 million.
Location: Mid-Atlantic corridor.
Status: Ending 190 positions.
Assessment: 60% probability. Significant financial pressure, strategic location.
University of North Texas (Denton, Texas)
Deficit: $45 million and growing.
President Harrison Keller: “Hard choices” ahead. Deficit is “structural, not just temporary.”
Location: Dallas-Fort Worth emerging data center market.
Assessment: 55% probability. Structural deficit plus market growth.
The Broader Crisis Creating Desperation
The dual enrollment crisis:
Universities face a two-factor enrollment collapse. The demographic cliff starts 2026. A 15-20 percent decline in college-age students over the next decade, rooted in 2008 birth rate drops.
But it’s worse than demographics alone. Students who DO exist are increasingly choosing not to attend. An NBC poll in November 2025 found 63 percent of voters say a four-year degree “isn’t worth the cost”. A dramatic reversal from a decade ago.
The driver: AI’s impact on entry-level employment. Computer science enrollment dropped 15 percent in Fall 2025, reversing a decade of growth. Entry-level hiring at the 15 biggest tech firms fell 25 percent from 2023 to 2024. Programmer employment collapsed 27.5 percent in two years. Entry-level jobs increasingly demand 2-3 years experience, creating a paradox where graduates can’t get the jobs they trained for.
Students are watching AI automate the “learning curve”. The tasks juniors used to perform to gain experience. Companies explicitly cite AI when cutting entry-level roles. Why pay for a four-year degree when AI is displacing the jobs that degree was supposed to access?
Universities aren’t just losing population. They’re losing trust in the value proposition itself
Federal funding cuts:
NIH: $18 billion reduction proposed.
NSF: $5.1 billion reduction.
Total federal research funding threatened: $60 billion.
Universities depend on overhead from federal grants.
State funding cuts:
Maryland cut $155 million. Multiple states reduced higher education budgets for FY2026. Universities forced to “make hard choices.”
International enrollment collapse:
Visa restrictions. Geopolitical uncertainty. Some universities are 50+ percent dependent on international tuition. That revenue stream is collapsing.
Commercial real estate debt crisis:
$930 billion in CRE loans mature in 2026. Universities hold significant real estate portfolios. Many need liquidity to refinance or avoid foreclosure.
These pressures create desperation. Desperation creates vulnerability. Vulnerability creates opportunities for buyers with cash.
TAXATION WITHOUT REPRESENTATION 2.0
In 1776, American colonists objected to British taxation without colonial representation in Parliament. “No taxation without representation” became a rallying cry.
In 2026, tech companies extract tax revenue through exemptions and deploy infrastructure without community representation in decisions. The parallel isn’t perfect. But it’s instructive.
The Modern Extraction
Virginia’s automatic tax exemptions:
$1.9 billion per year to data centers (2026). That’s 80 percent of ALL Virginia economic incentive spending. The exemptions are automatic “as of right”. No discretionary approval needed. Companies qualify by meeting thresholds: $150 million investment plus 50 jobs.
No public hearing on whether Amazon should receive $100+ million in tax exemptions over 15 years. The program exists. Amazon qualifies. Done.
Michigan’s grant mechanism:
$100 million to U of M/Los Alamos. Community resolutions ignored. Township powerless due to university zoning exemption. Only the state legislature can reverse the grant and that requires political will, committee hearings, votes, and likely gubernatorial approval.
The Power Imbalance
Traditional data center approval process:
- Developer proposes project to township
- Public hearings held
- Township board votes on approval
- Tax abatements negotiated publicly
- Community can organize opposition before approval
- Process is public even if flawed
University bypass mechanisms:
Virginia model:
- Private sale completed before community involvement
- Community learns after the fact
- Fights zoning requirements from weaker position (property already sold)
Michigan model:
- University bypasses local control through zoning exemption
- State grant approved without township input
- Community must convince state legislature to reverse
Both result in infrastructure deployment with minimal meaningful community input. Both shift leverage away from local control. Both create extraction without representation.
The colonists could at least see the taxes and protest them. Modern communities often learn about $427 million transactions after they close.
WHAT COMES NEXT
The 12-18 Month Test
We’ll know whether these are successful test cases within 12-18 months though the market may not wait for the results.
If U of M’s $100 million grant survives legislative challenge:
Expect more public university partnerships with federal entities. The template: Research mission plus zoning exemption plus state funding. The target: Public universities in data center corridors.
If Amazon successfully converts GW’s campus despite zoning requirements:
Expect more private acquisitions of distressed campuses. The template: Buy first, seek approval after from position of ownership. The target: Private universities with valuable land and deficits.
If both succeed:
Expect universities across the country to receive calls from data center developers with attractive offers for “non-strategic real estate holdings.” Community resistance mechanisms proven effective in 2025 may be rendered ineffective in 2026.
Questions Communities Should Ask Now
- Does our local university have satellite campuses in valuable locations?
- Is the university facing budget deficits or enrollment decline?
- Are we in or adjacent to data center corridors?
- Has the university mentioned “real estate monetization” or “non-strategic holdings”?
- What local authority exists to prevent private sales or public exemptions?
The Legislative Response
Michigan shows what’s possible:
HB 5362 to rescind the grant has 17 bipartisan co-sponsors. Senate Democrats introduced data center regulation bills. Multiple townships passed moratoriums. Communities organized before projects were approved.
Virginia shows the limits:
Sale announced after completion. County supervisor can “fight hard,” but the property is already sold. Zoning fight happens from a weaker position. The community gets input after the leverage has already shifted.
Which model prevails, in whole or in part matters for every community near universities and data infrastructure.
BOTTOM LINE
Two universities are converting educational assets to data infrastructure through fundamentally different mechanisms. One uses public exemption and state grants. The other uses private sale and automatic tax subsidies.
Both are happening simultaneously (2024-2026). Both face organized resistance. Both involve educational institutions under financial pressure selling or partnering their way out of deficits.
The evidence – timing, mechanism differentiation, replicability criteria – suggests these are test cases for scalable bypass strategies rather than isolated coincidences.
Brown University already announced plans to “monetize non-strategic real estate holdings.” The University of Chicago has a $160 million deficit in the nation’s largest data center market. The University of Maryland borders Virginia Data Center Alley with a $155 million budget shortfall.
Dozens more institutions meet the vulnerability criteria. The demographic cliff starts 2026. Federal funding cuts threaten $60 billion in research revenue. International enrollment is collapsing. $930 billion in commercial real estate debt matures this year.
Universities are desperate. Data center developers have cash. And two different bypass mechanisms are being tested right now to see which works better.
Michigan’s process is flawed but reversible through legislation. Virginia’s process doesn’t exist. The community learns after the sale closes. Both extract value from educational institutions and communities. One at least pretends to ask permission.
Is that a meaningful difference? Or just two routes to the same destination?
Communities will find out within 18 months. By then, the phone calls to university CFOs may have already been made. The “non-strategic real estate” identified. The deals negotiated under confidentiality provisions.
And by the time communities learn about it, the leverage may have already shifted.
SOURCES:
GW/Amazon Sale:
George Washington University sale announcement: https://president.gwu.edu/announcing-sale-gws-virginia-science-and-technology-campus
Washington Post (buyer confirmation): https://www.washingtonpost.com/education/2026/03/02/george-washington-university-ashburn-sale/
GW Hatchet (deed records, details): https://gwhatchet.com/2026/03/02/gw-sold-virginia-campus-for-427-million-to-amazon-for-data-center-development/
WTOP (Briskman quote, zoning details): https://wtop.com/dc/2026/03/gw-sells-ashburn-campus-with-few-details-about-sale/
U of M/Los Alamos:
Planet Detroit (U of M/Los Alamos, rescission bill): https://planetdetroit.org/2025/12/university-of-michigan-data-center-funding/
Crain’s Detroit Business (HB 5362, project details): https://www.crainsdetroit.com/politics-policy/bill-would-revoke-100m-university-michigan-ai-center
Bridge Michigan (Michigan moratoriums, legislative action): https://bridgemi.com/michigan-environment-watch/at-least-19-michigan-towns-pause-data-centers-no-one-knows-if-itll-work/
Federal Policy:
White House Executive Order (July 23, 2025): https://www.whitehouse.gov/presidential-actions/2025/07/accelerating-federal-permitting-of-data-center-infrastructure/
White House Fact Sheet (Executive Order details): https://www.whitehouse.gov/fact-sheets/2025/07/fact-sheet-president-donald-j-trump-accelerates-federal-permitting-of-data-center-infrastructure/
Davis Graham & Stubbs (federal lands implementation): https://davisgraham.com/news-events/the-trump-administrations-progress-to-site-data-centers-on-federal-lands-initial-steps-but-work-remains/
Tax Subsidies & Financial Data:
Virginia Economic Development Partnership (tax exemption details): https://www.vedp.org/incentive/data-center-retail-sales-use-tax-exemption
Good Jobs First (Virginia subsidy costs): https://goodjobsfirst.org/virginia-data-center-subsidy-costs-balloon-by-1051/
Higher Education Financial Crisis:
Pew Charitable Trusts (higher education financial crisis): https://www.pew.org/en/research-and-analysis/articles/2025/11/12/higher-educations-uncertain-fiscal-future
Inside Higher Ed (Brown real estate monetization quote): https://www.insidehighered.com/news/business/cost-cutting/2025/10/08/economic-uncertainty-spurred-campus-cuts-september
The Open Record L3C provides actionable intelligence for workers and communities navigating AI’s economic transformation.
theopenrecord.org
