The U.S. is in a high stakes race to lead in artificial intelligence and manufacturing, and energy dominance is key. AI’s explosive growth is driving unprecedented electricity demand, with projections suggesting U.S. data centers could consume 9% of the nation’s electricity by 2030, up from 4% in 2023. Meeting this demand requires reliable, scalable, and dispatchable power. Right now, natural gas is the only fuel that fits the bill.
A Trump-era focus on energy leadership is creating new opportunity. With abundant reserves, a resurgent manufacturing base, and policy tailwinds, the U.S. is positioning itself to power the AI revolution and revive its industrial sector.
How will the U.S. lead the AI and manufacturing boom?
The U.S. holds a favorable position at the intersection of AI and manufacturing. Energy dominance is at the heart of the nation’s competitive edge, but there are a couple of key factors at play:
1. Abundant, Reliable Energy from Oil & Gas
Right now, the U.S. leads globally in oil and gas production and exports. In 2023, the U.S. accounted for 15.6% of global oil production, surpassing both Russia and Saudi Arabia. Thanks to plentiful reserves in the Permian Basin and Bakken Formation, as well as advanced technologies like horizontal drilling, the nation has become an energy leader.
Since 2009, the country has been the largest natural gas producer in the world, thanks to productive shale formations like Marcellus, Utica, and Haynesville. Two years ago, America also became the top LNG exporter, passing up both Qatar and Australia.
2. Investments in Manufacturing
The same boom that turned the U.S. into a net energy exporter is now fueling a manufacturing resurgence. With natural gas abundant and prices low, energy intensive industries (from semiconductors to steel) are expanding domestic production.
U.S. firms are investing heavily in gas turbine manufacturing and other energy infrastructure to support AI data center builds. GE Vernova’s Schenectady facility is receiving $600 million to revive U.S. manufacturing capacity for turbines that will supply data centers. In Pennsylvania, the company is investing $100 million in technologies to help modernize the grid.
3. Spiking Demand from AI
AI data centers are not only energy intensive, but require always on, high capacity power with near perfect uptime. AI-driven power demand could grow thirtyfold by 2035, up to 123 GW, a surge that something like intermittent renewables alone cannot meet. Dispatchable and abundant natural gas, which already supplies over 40% of U.S. electricity, can fill the gap. Today, new power plants and high-efficiency turbines are being deployed near data hubs to minimize transmission losses.
4. AI & Manufacturing Synergies
AI is driving a revival of domestic semiconductor, hardware, and infrastructure production, supported in part by Trump-era industrial policies favoring reshoring and fossil-based energy projects. AI is also making manufacturing more efficient. Predictive maintenance, smart sensors, and industrial IoT are slashing downtime in factories and driving productivity gains.
It’s a cycle. Cheap energy attracts manufacturing, AI optimizes production, and reshoring strengthens supply chains. This all reinforces U.S. competitiveness.
5. A Favorable Political Climate
Recognizing the stakes, Washington is moving aggressively to clear roadblocks for energy infrastructure. The Trump administration has released a series of executive orders this year, some of which aim to:
– Fast-track permitting for data centers and related energy or manufacturing projects
– Allocate federal land for data centerswith priority grid connections
– End subsidies for foreign controlled energy sources to level the playing field for domestic production
– Assert energy dominance by cutting red tape, enhancing private sector investments, and promoting innovation
Critics warn that undercutting clean energy progress could jeopardize long-term competitiveness in AI and manufacturing, given global momentum in renewable technologies built by rivals like China. But the fact remains that we need natural gas here and now to keep the lights on.
The Path Forward
The U.S. is uniquely positioned to lead the AI race, thanks to its energy abundance, industrial capacity, and policy flexibility. But maintaining this edge will require more than just doubling down on natural gas.
In the short term, natural gas remains the only scalable solution that keeps data centers running without interruption. In the long term, advanced nuclear, next-gen tech, and a smarter grid will determine whether America sustains its lead or cedes ground to competitors.
AI leadership will be limited by energy. And right now, the U.S. seems to have what it takes.
Electric Deals of the Month
Google to Spend $25 Billion on Data Centers and AI Infrastructure in PJM
PJM is the country’s largest grid, in charge of power for 13 states across the mid-Atlantic, Midwest and South. The region has been struggling to meet growing power demand. To address this problem, Google is also planning to spend $3 billion to modernize two Pennsylvania hydroelectric plants. This is part of a larger agreement between Google and Brookfield Asset Management to purchase 3,000 MW of hydroelectric power.
Texas to Require Data Centers to Pay for Interconnection Costs
Texas has just passed legislation requiring large loads like data centers to help pay for interconnection costs. The bill applies to customers pulling 75 MW or more from the ERCOT grid. Additionally, the bill affects facilities with 50% or more of onsite generation. If grid capacity experiences shortfalls, these facilities may be required to give up some power to the grid or curtail their own loads.
Hyperscalers Ramping Up Large Lease Data Centers Again
According to analysts at TD Cowen, hyperscalers like Amazon and Microsoft are now showing greater interest in leasing at scale. It was previously reported that Microsoft had canceled up to 2GW in projects, since they were no longer the only compute provider for OpenAI. However, analysts are now saying that not all 2GW were cancelled, only delayed 2 years. Amazon too seems to be increasing spend, supposedly having recently secured a 1GW+ customer. This follows reported delays in data center leases earlier this year.
For more Electric Buzz, keep an eye out for next month’s newsletter!
