The US Air Force is preparing to lease underutilized land on its installations to private companies for the construction of artificial intelligence (AI) data centers. The initiative aims to bolster national security infrastructure while advancing energy resilience and digital modernization goals. The move reflects a growing intersection between military basing strategy and the demands of high-performance computing in defense applications.
Strategic Use of Underutilized Base Land
According to recent statements from Ravi Chaudhary, Assistant Secretary of the Air Force for Energy, Installations and Environment, the service plans to leverage “non-excess” land—areas not actively used but not surplus—to host commercial data centers. These sites will be leased under enhanced use lease (EUL) authorities that allow military departments to partner with private entities while retaining ownership of the land.
The Air Force has already identified at least a dozen potential locations across its domestic base network. While specific sites have not been publicly disclosed due to security considerations and ongoing negotiations, Chaudhary noted that these areas are being selected based on proximity to fiber-optic infrastructure, power availability (including renewable sources), and alignment with mission needs.
This approach enables dual-use benefits: it supports national security by increasing domestic data processing capacity—vital for AI model training and real-time analytics—while also improving base resiliency through potential microgrid integration or energy-sharing agreements with tenants.
AI Infrastructure as a National Security Asset
The Department of Defense (DoD) has increasingly emphasized the importance of AI in future warfare domains. From autonomous systems and ISR fusion to predictive maintenance and cyber defense, AI workloads require massive computational resources. However, most current DoD infrastructure is not optimized for such demands.
By inviting commercial hyperscale or modular data center operators onto Air Force installations, the service can facilitate secure processing environments close to mission operations without bearing full capital costs. These facilities could support both government-only workloads (via secure enclaves or air-gapped segments) as well as general-purpose compute aligned with federal priorities like the Joint Artificial Intelligence Center (JAIC), Project Maven, or CDAO initiatives.
This model mirrors broader federal efforts such as Executive Order 14110 (October 2023), which mandates responsible AI development while encouraging public-private partnerships in critical infrastructure domains.
Energy Resilience Through Co-Located Infrastructure
A key enabler of this initiative is energy resilience—a growing concern amid rising cyber threats to grid infrastructure and climate-induced instability. Data centers are power-intensive facilities often requiring tens of megawatts per site. By co-locating them on bases with existing or planned renewable energy projects (solar arrays, battery storage), both parties can benefit from resilient microgrids or behind-the-meter generation models.
Chaudhary emphasized that these projects will be designed “with resilience baked in,” potentially including black-start capabilities or islandable operations during grid outages. This aligns with broader DoD goals under its Climate Adaptation Plan and Installation Energy Plans (IEPs).
In some cases, tenants may invest directly in base energy infrastructure as part of their lease agreements—offsetting military utility costs while ensuring uptime for their own compute operations.
Enhanced Use Lease Mechanism Enables Flexibility
The Enhanced Use Lease authority allows military services to lease non-excess property for up to 50 years in exchange for cash or in-kind consideration that benefits installation missions. This mechanism has previously been used for commercial housing developments, solar farms, logistics hubs—and now potentially hyperscale compute nodes.
Unlike traditional base realignment processes that require Congressional approval or BRAC authorization, EULs offer a more agile path for leveraging existing assets without divestment. As long as core mission functions are not impaired—and environmental compliance is met—the Air Force can proceed through internal review processes led by AFCEC (Air Force Civil Engineer Center).
EULs also allow flexibility in structuring deals around evolving tech requirements; e.g., modular container-based data centers vs fixed buildings; hybrid cooling systems; edge compute vs centralized storage; etc.
Challenges Ahead: Security Integration & Community Impact
While promising in concept, several challenges remain before these AI data center projects become operational:
- Security Integration: Ensuring physical and cyber separation between commercial tenants and sensitive USAF operations will require robust zoning protocols and possibly SCIF-rated buildouts if classified workloads are involved.
- Zoning & Environmental Compliance: Some bases may face NEPA hurdles or local opposition depending on noise levels from HVAC systems or visual impact near residential areas.
- Civic Coordination: Local governments may seek assurances about job creation benefits or traffic mitigation if large-scale construction occurs near urban zones adjacent to bases like Travis AFB or Joint Base San Antonio.
The Air Force has indicated it will work closely with local stakeholders through its Community Partnership Program offices at each installation where EULs are pursued.
A Model for Future Defense-Tech Synergies?
This initiative could serve as a template across other services—e.g., Army Futures Command’s interest in edge computing nodes at logistics hubs—or even NATO allies seeking sovereign cloud capabilities within secure perimeters. With global demand for AI compute outstripping supply chains for GPUs and power-dense real estate near fiber backbones becoming scarce, military installations offer unique advantages: controlled access zones; hardened utilities; existing command structures; and strategic geography away from coastal risk zones.
If successful at scale, this model could evolve into a new class of dual-use defense infrastructure—where bits meet bases—not just bullets meet bunkers.
