Universities trade campus land for AI data centers amid financial pressures

The rapid expansion of artificial-intelligence infrastructure is reshaping how U.S. universities manage their real-estate assets. George Washington University announced the sale of its roughly 120-acre Virginia Science and Technology Campus in Ashburn, Virginia, to Amazon Data Services for $427 million; the university will retain use of the property for up to five years, allocate a large share of the proceeds to a new endowment, and provide modest one-time bonuses to eligible faculty and staff.

Ellen Scully-Russ, a former GW professor, described the transaction as part of a larger "land grab" by data-center operators and criticized the absence of prior consultation with faculty or students. Financial strains such as reduced federal research funding and shifts in international student enrollment have pressured GW to pursue the deal. In a different strategy, the University of Michigan and Los Alamos National Laboratory are moving forward with a $1.25 billion high-performance computing and research center in Ypsilanti Township, intended for scientific workloads rather than commercial cloud services, though the Ypsilanti Township Board of Trustees has formally opposed the project. Both cases highlight the tension between academic missions and the lucrative demand for AI-related infrastructure.