America’s AI future will be powered by private capital, or not at all

The Energy Information Administration now forecasts the strongest four-year run-up in U.S. electricity demand since 2000, driven largely by data centers and large computing facilities. By 2030, those facilities could account for up to 17% of all U.S. electricity consumption. AI loads do not behave like the demand growth we managed in the past. They are continuous, regionally concentrated, and intolerant of interruption. A data center campus does not throttle back at night, nor can it simply power down when demand spikes or temperatures rise.

Some will tell you the question is whether to invest. That is the wrong question. The grid built over the past century was not designed for the future America needs to thrive, and pretending otherwise is not a strategy. The real question is whether the country will modernize its energy infrastructure proactively or be forced to do so reactively under greater economic and political pressure.

Without serious new capital, here is what we get: deferred maintenance, aging infrastructure, tightening reserve margins, and the rate shocks that follow. Those costs do not fall on Wall Street. They fall on families and small businesses, the least able to absorb them. With serious new capital, here is what we get: a more diversified, more resilient grid; modern generation and transmission built for continuous loads; and domestic capacity to keep AI and advanced manufacturing in the U.S., rather than ceding the next industrial era to Beijing.

Consider one example. AES, a Fortune 500 utility serving customers in Indiana and Ohio, publicly acknowledged it needed outside investment to support growth beyond 2027. Without new capital, the company itself signaled it might have to cut or eliminate its dividend, a clear sign of a balance sheet too thin to expand the grid without passing price shocks to customers. In March, a consortium led by Global Infrastructure Partners and EQT AB announced a $33.4 billion agreement to acquire AES and provide that capital. The result is a utility with the financial flexibility to build the modern generation, storage, and transmission its customers will require. That is what the energy dominance agenda looks like in practice.

Transactions such as this are exactly what Trump’s $500 billion Stargate initiative, announced in January 2025, will require to succeed. Stargate is the most ambitious AI infrastructure commitment in American history. It will not be financed by the federal government, and it should not be. The capital, the operational expertise, and the willingness to take on project risk are sitting in the private sector.

Policymakers do not need to pick winners and losers. They need to make sure capital can flow to projects that strengthen reliability and serve American consumers, with transparent pricing, clear market signals, and a regulatory environment that does not punish the partners willing to put real money at real risk.

Natural gas, nuclear, and renewables each play distinct roles. The honest case for private capital is that it makes a balanced portfolio possible at the speed and scale that demand now requires. Energy policy has always meant balancing cost, reliability, security, and stewardship. What is different today is the scale and urgency of demand.

We have met transformational energy moments before, and we can do it again. Trump has set the direction. The capital is ready. The only question is whether Washington will get out of the way.