Already, the electrical grid in the US is groaning under the strain of explosive demand for power driven by data centers clamoring to ride the tsunami of AI adoption. AI’s appetite for energy is voracious, and the Trump administration wants to feed it.
Legislation and strategies from the federal government aim to keep the US in the lead of a fevered global race. They will play a significant role in shaping the way energy infrastructure will keep up, or buckle.
The One Big Beautiful Bill Act unleashes millions of dollars of funding for the federal government’s AI efforts, and it makes big changes to renewable energy incentives. The AI Action plan aims to speed the construction of data centers and power generation infrastructure.
CIOs and other enterprise technology leaders are well into the AI adoption race, but big questions loom about energy demands and sustainability. As AI and its insatiable energy needs ramp up, how will these leaders think about federal policies, the business needs of their organizations, and long-term sustainability?
A Shift Away from Wind and Solar Power
The reconciliation law does not entirely swerve away from low-carbon energy; geothermal and nuclear power maintain many incentives, according to the Council on Foreign Relations. But tax incentives for solar and wind power introduced with the Inflation Reduction Act (IRA) of 2022 shrink or vanish entirely under the Big Beautiful Bill.
Power-hungry data centers and their customers are likely to look elsewhere if wind and solar power projects slow or stop entirely.
“I am seeing data center operators scrambling to do everything they can to turn on data centers, getting power from anywhere they can,” says Jason Eichenholz, founder and CEO of Relativity Networks, a company that develops hollow core fiber technology. “The Big Beautiful Bill or not, they want the power.”
More Data Centers and Power, Faster
The second pillar of the AI Action Plan is dedicated to building out an American AI infrastructure. It recommends policy actions that would make it easier and faster to construct data centers, semiconductor manufacturing facilities and energy infrastructure. The plan recommends reducing or eliminating environmental regulations under the Clean Air Act and Clean Water Act, among others. It also puts forth the idea of making Federal lands available for construction.
It also emphasizes the need to develop a power grid to handle AI’s demands. The plan calls for “…new energy generation sources at the technological frontier (e.g., enhanced geothermal, nuclear fission, and nuclear fusion).”
An Uncertain Future
Increasing energy costs are an unsurprising consequence in a landscape of demand that outweighs capacity. It is already happening. Reduced investment in energy sources like wind and solar could contribute to those rising costs.
“Without taking advantage of a mature energy generation technology like solar or wind, then if you take that off the table … the other sources of energy are going to have to keep up with the growing demand, which will increase the prices,” says Gregg Semler, founder and CEO of InPipe Energy, a renewable energy company.
It seems inevitable that cracks in the energy infrastructure are going to start to show. “We’re going to start experiencing not only price increases, but the fragility of the grid as well,” says Semler.
Competition for power isn’t the only sustainability concern. Data centers require massive amounts of water and land. Bloomberg found that approximately two-thirds of data centers built or in development since 2022 are located in areas experiencing high levels of water stress. Mark Zuckerberg announced plans for data centers with mythical names, Prometheus and Hyperion, which will take up very real, very large areas. “Just one of these covers a significant part of the footprint of Manhattan,” Zuckerberg declared on Threads.
The communities that play host to data centers, willingly or not, are feeling the impact of the mad rush to provide AI with the resources it needs.
How do CIOs and other enterprise leaders think about the sustainability issues that come with the AI and data center boom? It appears that sustainability hasn’t completely fallen off the radar, but it isn’t driving the bulk of decision making. A survey conducted by Seagate, a mass-capacity data storage company, found that 95% of respondents are worried about the environmental impact of data centers. Yet just 3.3% of them report prioritizing those concerns when making purchasing decisions.
“ESG and sustainability are 100% on the list. Right now, it’s not No. 1 on the list. Speed to interconnection and speed to power is number one on the list,” says Pete DiSanto, executive vice president, data centers at Enchanted Rock, a renewable energy and power solutions company.
Whether the mismatch between demand and available power is a short-term bottleneck or a long-term pitfall is up for debate.
“The bottle neck right now is everybody wants firm power,” says DiSanto.
That craving for power is going to force everyone–data center operators, utilities, and big tech–to get creative. The need for diversified energy sources is clear.
“We’re trying to diversify it, but when you take solar and wind off, you’ve gone backwards in terms of diversification,” says Semler.
The Trump administration is pushing for more natural gas, oil, and coal production, along with “…hydropower, biofuels, critical mineral, and nuclear energy resources,” according to the Unleashing American Energy executive order.
That represents a mix of renewable energy sources and energy sources that make significant contributions to greenhouse gas emissions. If data center operators and their customers want to prioritize sustainability, they are going to have to lean into the renewable options.
“They have to look for things like geothermal energy, hydropower, nuclear power, just other sources of energy besides coal and natural gas,” says Semler. “But I don’t think they’re really going to care that much.”
Money is certainly being poured into different energy sources. Meta is involved in a geothermal energy project. Amazon, Google, and Microsoft are investing in nuclear power. Google also has a $3 billion hydropower deal.
“If the utilities and regulators can’t find a way to deliver … gigawatts of power, big tech will figure out how to do it because that’s what they do,” says Eichenholz.
For the time being, sustainability has not been necessarily forgotten, but it does seem to have taken a back seat, particularly from a policy perspective.
“AI is just such an energy hog that they’re going to scoop up any kilowatts that they possibly can at whatever cost, whatever is producing them,” says Semler.
