Globally, the International Energy Agency (IEA) estimates that electricity demand from data centers will more than double by 2030 to about 945TWh, and that in the United States, they may account for nearly half of all electricity-demand growth by that date.

The IEA reports that electricity use from data centers has grown by about 12 percent annually since 2019, outpacing total global electricity demand growth by a factor of four. The United States remains the largest single market for data center electricity use, with per-capita consumption projected to rise from around 540 kilowatt-hours (kWh) in 2024 to more than 1,200 kWh by 2030. This figure is equivalent to ten percent of the annual electricity consumption of an average US household.

The physical footprint of data centers, including land, power, cooling, and connectivity, means they face challenges similar to those of traditional infrastructure systems. Treating them as such would align policy, regulation, and investment with broader infrastructure priorities such as power, water, and telecommunications.

The IEA also notes that data center development is increasingly concentrated in a few states and regions, which can place disproportionate pressure on local infrastructure and environmental resources. By treating data centers as infrastructure, policymakers would align support, regulation, and integration with the broader systems of roads, power, water, and telecommunications.

Expanding data center capacity places heavy demands on existing infrastructure systems, especially power delivery, cooling (often water-intensive), real estate, and high-capacity network connectivity. Projected 2028 loads of 325 to 580TWh correspond to 74 to 132GW of demand. The World Resources Institute warns that unmanaged load growth could overwhelm current grid planning, leading to higher consumer bills and lower reliability.

Roughly half of data center electricity is consumed by IT equipment, while the rest is used to power cooling systems. McKinsey observes that US states will need to plan strategically for the rapid expansion of data centers, balancing economic opportunities with local constraints on power, water, land, and labor.

Consequently, states capable of effectively managing these trade-offs through infrastructure investment, community engagement, and coordinated planning can unlock billions of dollars in direct and indirect economic growth.

The Congressional Research Service likewise highlights that no binding federal standards currently govern private-sector data center energy use, and that limited, fragmented data collection leaves policymakers with an incomplete picture of the sector’s resource demands. Strengthening coordination between federal and state authorities could improve oversight, align energy and infrastructure planning, and ensure that data center expansion supports national efficiency and sustainability goals.

The digital economy now runs on data centers. They enable cloud computing, AI training, streaming, e-commerce, telehealth, remote learning, and other digital services that sustain both public and private operations.

The IEA notes that electricity consumption from data centers has grown rapidly in recent years and will more than double by 2030, with significant implications for power systems and energy policy. Outages in power, cooling, or connectivity can disrupt activities across finance, health, logistics, and government. Given their systemic role, treating data centers as core infrastructure comparable to highways or ports could help align policy and investment with their economic and national-security significance.

Inclusion in infrastructure legislation would also enable coordinated planning for resilience through reliable power supply, secure fiber connectivity, disaster-risk management, and cybersecurity.

Expanding data center capacity raises pressing environmental challenges. US data centers generated over 105 million tons of CO₂e in 2023, accounting for about 2.18 percent of national emissions, with emission intensity 48 percent higher than the US average.

Water use for cooling, land requirements, and heat discharge into the surrounding environment also create localized ecological pressures. The World Resources Institute warns that unmanaged load growth could unfairly shift grid-upgrade costs to consumers if cost-allocation mechanisms are incorrectly designed.

Policy design should therefore integrate sustainability into infrastructure planning. This includes renewable-energy procurement targets, efficiency standards such as PUE, and water-management provisions to protect scarce regional resources.

An infrastructure bill provides an effective policy framework for data center development because it links energy, broadband, land, and cooling systems under a coordinated investment approach rather than through isolated funding. Most data center capital is private, but the supporting systems, such as electric grids, fiber networks, and land-use planning, are public responsibilities.

At the outset, the inefficiency of state-level incentives for data centers is stark: in Central Ohio, state and local governments forfeited more than $750 million in tax revenue for fewer than 800 permanent jobs, which means some projects cost over $2 million in tax breaks per long-term position.

This illustrates how fragmented incentives have led to diminishing returns and fiscal inefficiencies, and these uncoordinated incentive programs have also strained public infrastructure: data center power demand in Ohio alone is projected to rise 700 percent by 2030, which is driving billions in grid investment that utilities are passing on to ratepayers.

Because data centers depend on shared energy and broadband networks, an infrastructure framework allows joint planning among utilities, technology firms, and local governments. This mirrors prior federal designations of broadband and energy transmission as critical infrastructure, which successfully leveraged blended finance and federal–state coordination.

Data centers have become essential components of national infrastructure, with evidence showing they could account for roughly 12 percent of US electricity consumption by 2028 as AI workloads expand. Yet this growth places increasing pressure on power grids, cooling systems, broadband networks, and land resources.

Including data centers in the following infrastructure bill would align public policy with twenty-first-century realities, supporting digital competitiveness, improving resource efficiency, and embedding sustainability and resilience across regions. This would bridge the digital–physical divide and ensure that the coming decade of AI-driven innovation rests on an efficient and secure foundation.