A university that once housed classrooms and research labs is about to become another node in America’s data center map. George Washington University confirmed in February that it sold its Virginia Science and Technology Campus to Amazon Data Services for $427 million, a deal that puts a very specific price tag on the growing tension between AI infrastructure and universities across the country. The sale, along with a separate $1.25 billion computing project the University of Michigan is pursuing in Ypsilanti Township, shows how campuses are increasingly being treated as real estate assets in a race for land, power, and fiber connectivity.
Summary
Key takeaways
- George Washington University sold its 120-acre Virginia Science and Technology Campus to Amazon Data Services for $427 million.
- GW can keep using the site for up to five years while it relocates programs, and proceeds will fund a new endowment plus one-time faculty bonuses of $2,500 for full-time and $1,250 for part-time staff.
- A former GW professor says the sale happened with little transparency and reflects a broader erosion of higher education resources.
- The University of Michigan is planning a $1.25 billion high-performance computing center with Los Alamos National Laboratory, aimed at scientific research rather than commercial cloud services.
- The Ypsilanti Township Board of Trustees unanimously opposed the Michigan project, signaling local resistance to AI infrastructure expansion.
George Washington University’s $427 Million Sale to Amazon Data Services
The short answer is simple: GW needed money, and Amazon needed land. The university’s roughly 120-acre campus in Ashburn, Virginia sits inside Loudoun County’s so-called Data Center Alley, one of the densest concentrations of data centers on Earth, which made the property valuable for reasons that had nothing to do with academics. When GW first announced the sale, it declined to name the buyer or the price, citing a confidentiality agreement. That didn’t last long. A student newspaper tracked down both details within a day, according to former GW professor Ellen Scully-Russ, who taught at the Ashburn campus starting in 2009.
Financial pressures behind the sale
GW has been squeezed from multiple directions. The university pointed to declining federal research funding and shifting policies affecting international students as part of the financial backdrop that made the sale necessary. Scully-Russ said those pressures made the decision easier to understand, even if she disagreed with how it played out. She also described a slower decline at the Ashburn campus that predated the sale itself, calling it a “drip, drip, drip” of lost faculty positions and eliminated teaching assistant roles over five or six years.
Terms allowing continued university use
The deal isn’t an immediate eviction. Under the agreement, GW can continue operating on the property for up to five years while it figures out how to relocate the programs and offices that were based there. That transition window gives the university breathing room, but it also puts a clock on decisions about where displaced academic programs will land next.
Allocation of sale proceeds
GW says a significant share of the $427 million will go into a new endowment meant to strengthen its long-term finances. The university also set aside money for staff: eligible full-time faculty will receive a one-time bonus of $2,500, while part-time faculty will get $1,250. GW has framed the sale as a strategic move to protect its academic mission, not a retreat from it.
Academic Concerns and Criticism of the Sale
Not everyone at GW sees the transaction as good news. Scully-Russ, who spent years teaching at the Ashburn campus, argues the deal was handled without input from the people it affected most. “There was no transparency,” she told Fortune, adding that neither faculty nor students were told a sale was even under discussion before it happened.
Her larger worry is about what the sale represents for higher education broadly. “We’re chipping away at our higher ed resources to make room for this technology to expand without any accountability,” she said, “or any other kind of thoughtful progression, which is really a concern for me.” That framing captures a real tension: as AI infrastructure and universities compete for the same land, campuses risk losing space that once served research and teaching to server racks and cooling systems instead.
Scully-Russ is also skeptical that the trade-off will pay off long term. She described the current wave of data center construction as “just a short-term bubble,” warning that a single technological leap could leave today’s massive facilities obsolete. “Then what happens?” she asked — a question that goes to the heart of whether universities selling land now are making a durable financial move or simply cashing in during a temporary surge in demand.
University of Michigan’s $1.25 Billion AI Computing Center and Local Opposition
Not every university is selling land to data center operators — some are building their own. The University of Michigan, working with Los Alamos National Laboratory, is planning a $1.25 billion high-performance computing center in Ypsilanti Township designed to handle heavy computational workloads for scientific research.
Focus on scientific AI research rather than commercial use
Michigan is careful about the label it uses. The university avoids calling the facility a “data center,” preferring “computing center” instead, and says its computational output won’t be used for cloud computing, streaming, e-commerce, or other consumer and business applications. Instead, the school says the center will support research in medicine, materials science, clean energy, engineering, and national security. That research still demands enormous amounts of electricity, however: the facility is projected to start at about 50 megawatts of power consumption and eventually reach 100 megawatts.
Community pushback from Ypsilanti Township Board
Local officials aren’t convinced the distinction matters. The Ypsilanti Township Board of Trustees unanimously passed a resolution opposing the project, declaring what it called “strong and unequivocal opposition.” The University of Michigan did not immediately respond to a request for comment on the resolution. The pushback shows that even when a university frames a project as academic rather than commercial, communities near proposed sites are treating it the same way they would treat any large-scale data center: as a facility that will draw heavily on local power and water resources.
Broader Impact of AI Infrastructure Demand on Universities
Both cases point to the same underlying force. Data centers need large parcels of land with reliable access to electricity, water, and fiber connectivity, and universities frequently happen to own exactly that kind of property. That’s turning campuses into targets, whether the university is selling outright, like GW, or building its own facility, like Michigan.
Why does this matter beyond these two schools? Because it signals a pattern that could spread to other institutions facing similar budget strain. As federal research funding tightens and enrollment pressures mount, more universities may find that their land is worth more as AI infrastructure than as classroom or lab space — a trade-off that raises real questions about transparency, community input, and whether the technology driving these deals will still be relevant by the time proceeds are fully spent. Scully-Russ’s warning about a “short-term bubble” adds another layer: if today’s massive computing facilities become obsolete after the next technological leap, universities that sold land to fund one-time bonuses and endowments may be left without either the property or a lasting financial cushion.
FAQ
Why did George Washington University sell its Virginia Science and Technology Campus?
The sale was motivated by financial challenges including declining federal research funding, changes affecting international students, and broader financial pressures facing higher education institutions.
What are the terms of George Washington University’s sale of its Virginia campus?
GW sold the property to Amazon Data Services for $427 million but can continue using it for up to five years while relocating programs and offices.
What is the focus of the University of Michigan’s planned computing center?
The center is designed for scientific research applications of AI, supporting fields like medicine, materials science, clean energy, engineering, and national security, rather than commercial cloud computing.
Why is there opposition to the University of Michigan’s computing center project?
The Ypsilanti Township Board of Trustees unanimously opposed the project, reflecting local community concerns over the facility’s scale and resource demands.
Article produced with the assistance of artificial intelligence and reviewed by the editorial team.
