A recent request for information from the Department of Energy seeks ambitious, creative, and scalable solutions to data center infrastructure sited on DOE lands with colocated power. It needs to prioritize those solutions by speed-to-power. Vice President JD Vance recently affirmed that “there is no AI future without energy security and energy dominance,” and the United States must act quickly to secure its competitive advantage in AI dominance through unbridled energy dominance. This recent Request for Information on the siting of AI infrastructure on DOE lands is an example of the Department of Energy setting out to do just that. In our response, C3 Solutions argued that speed-to-power should be the chief consideration when assessing the projects and that a variety of tools should be used to achieve speed-to-power.

Artificial intelligence has emerged as a critical component of American national security and offers paths toward profound abundance and prosperity. It is right and good that President Trump and Secretary Wright have focused national attention and resources on building the infrastructure and culture necessary to even further accelerate American leadership in artificial intelligence. This request for information is a welcome call to utilize our great National Laboratories to support artificial intelligence infrastructure. 

Our response centered on the view that, for the effective advancement of American artificial intelligence leadership, the chief, if not all-surpassing, consideration should be speed-to-power—the time in which a data center moves from first approval to fully powered and operational. This is an especially important consideration in this process where massive investments will be made in federally owned lands and infrastructure. By adopting speed-to-power as the core objective of the deployment phase of this initiative, projects will be incentivized to get data centers powered and operational as quickly as possible, reducing downtime and providing opportunity for tremendous gains in artificial intelligence research and training. Other priorities should be subsidiary to speed-to-power in order to maximize the effectiveness of the projects which are being supported and sited under this initiative. 

A key way of prioritizing speed-to-power as the metric for these projects is through milestone-based awards for achieving certain power generation goals within a specified timeline. A tentative framework could be a set of five milestones ranging from six months to initial operation with at least 5 MW of power to five to ten years for fully scaled multi-gigawatt generation exclusively for AI data centers. This vision for multi-gigawatt-scale data centers follows suggestions from other scholars. The DOE parcels should be used to establish Special Compute Zones—federally designated areas where DOE is granted preemptive siting and permitting authority—to concentrate and accelerate the deployment of co-located AI and energy infrastructure. 

Finally, the Department should utilize every legally available means to streamline and speed up this process. In particular, DOE should make use of Other Transaction (OT) and Defense Production Act (DPA) authorities to speed up the acquisition of technology and agreements. DOE has historically not used OTs as widely or as frequently as other agencies, but this can change without any Congressional action. Using OTs for the siting of AI infrastructure is sensible and urgent, and could help to begin to usher in a sea-change at DOE where OTs are used much more widely. Further, because President Trump has declared a national energy emergency, the DPA can be used to fast-track supply chain commitments and investments for the infrastructure of these projects. These non-traditional authorities are essential to building AI infrastructure on DOE lands as quickly as possible. 

In conclusion, securing America’s AI advantage hinges on making speed-to-power the overriding priority for infrastructure development on DOE lands. This means selecting power-ready sites with existing energy footprints, grid assets, and completed environmental reviews—while simultaneously enacting bold regulatory reforms through federally designated Special Compute Zones. Within these zones, DOE should be granted preemptive, time-limited permitting authority (e.g., 6-month deadlines with a presumption of approval), empowered to apply Categorical Exclusions under NEPA using Defense Production Act (DPA) authorities, and backed by statutory protections against post-approval legal delays. Execution must be anchored in aggressive, technology-neutral deployment milestones—with meaningful financial incentives for early delivery and penalties for delay—favoring firm power sources such as gas, nuclear, and geothermal. To ensure delivery is feasible, DOE must also confront energy supply chain bottlenecks head-on through DPA-rated procurement and targeted investments to expand secure domestic manufacturing of long-lead components.

The views and opinions expressed are those of the author’s and do not necessarily reflect the official policy or position of C3.