U.S. May Need $511 Billion in New Power Capacity for AI Growth

The next phase of artificial-intelligence investment may hinge on electricity generation and grid access as much as on chips, servers and cloud platforms, with the United States potentially requiring about $511 billion in new power-generation capacity by 2030.

The estimate assumes U.S. electricity demand grows at a 3% annual compound rate. It excludes transmission and distribution investments needed to deliver power to data centers.

The projected buildout reflects a wider infrastructure constraint around AI. Additional computing capacity depends on power generators, utilities, substations, transformers, grid equipment, permitting and suitable data-center sites.

U.S. interconnection queues held about 2.6 terawatts of proposed generation and storage capacity at the end of 2023. Projects that entered commercial operation that year had waited a median of about five years to connect to the grid.

That timing pressure is also influencing state policy. New York Gov. Kathy Hochul signed an executive order on July 14, 2026, creating a one-year pause on new hyperscale data centers while the state develops new standards.

“The bottom line is that progress shouldn’t arrive with a higher utility bill, deleted water supply or noise pollution,” Hochul said.

The pause highlights the local effects of data-center expansion, including pressure on electricity systems, water supplies and communities. For AI developers, the availability of land and computing equipment may not be enough if projects cannot secure power connections.

The $511 billion estimate covers new generation capacity only. Transmission and distribution networks, which move electricity from power plants to data centers, would require additional investment beyond that figure.