The artificial intelligence arms race is here, and it runs on electricity. If even a fraction of Silicon Valley’s AI hype proves true, America should be moving mountains for the power necessary to fuel that race — especially with China now hot on our heels after DeepSeek’s recent breakthroughs. Unfortunately, too much red tape, uncertainty, and ideology are in our way.
AI development demands enormous power as we face historical growth in energy requirements from new manufacturing and reindustrialization. Never has an “all of the above” energy strategy been more urgent, but culture wars over different energy sources, “Not in my Backyard” vs. “Yes, in my Backyard” battles over permitting, and sobering supply chain realities are stymieing our efforts to meet that demand.
Hyperscalers and policymakers alike have been banking on a default “if all else fails” plan for natural gas to power us today and assuming advanced nuclear will save us tomorrow. But while these will be critical tools for our energy future, we need to be clear-eyed about their challenges over the required timelines. Natural gas turbines now have a three-year waitlist, while the first small modular nuclear reactors are not expected until the 2030s. With many companies expecting artificial general intelligence featuring human-like capabilities to emerge during this presidential term, those timelines simply won’t do.
The good news is that no physics breakthrough is required to solve this problem. We need no polluting sacrifices or exotic technologies. We just need to get out of our own way and permit ourselves to build, and Texas may be showing us the way forward.
While nationwide electricity production has plateaued for decades, leaving us flat-footed for this moment of growth, Texas has expanded its power-generating facilities by a whopping 35% in the last four years alone. The state’s energy ecosystem is powering a booming economy, strong population growth, high-tech innovation ecosystems, and advanced manufacturing. Its open energy market and flexible permitting frameworks make it easy to build, and the results are hard to argue whether you’re an oilman, a tech giant, or an environmentalist.
Texas leads the country in clean energy deployment, and on sunny days, it can even get most of its electricity from solar alone. It is also lapping the country in advanced battery technologies, buttressing grid resilience against Winter Storm Uri-style disruptions while turning surplus midday solar energy into firm, dispatchable nighttime power. That energy abundance is powering huge new investments in manufacturing and data centers while even driving the electrification of oil and gas infrastructure, making drilling cheaper and cleaner than ever.
Congress and the administration should act to help the rest of the country get out of its own way, become more like Texas, and realize the energy dominance required for this moment.
First, building must become easier. Permitting reform will support energy deployment generally and overdue transmission investments especially. Moving surplus power more effectively — such as midday solar in southwestern states — to where it’s needed around the country will make our grid more secure, resilient, and prepared for our future of electricity abundance.
Second, we must give industry the certainty to invest in America’s energy future. Lawmakers should protect tools that already support our energy dominance, such as the technology-neutral 45E/Y tax credits that power next-generation nuclear and geothermal industries. They should also support the related 45X incentives to reshore the manufacturing of batteries and other critical energy technologies.
THE US DESPERATELY NEEDS MORE ARTIFICIAL INTELLIGENCE DATA CENTERS
Finally, we need the administration to turn its Department of Government Efficiency chainsaws toward red tape. That includes working with industry to untangle natural gas supply chains and streamline nuclear construction while unleashing all the shovel-ready projects currently in line to connect to our nation’s power grid, 95% of which are already zero-carbon. That should include expediting national and local approvals for projects on federal lands ready to support our AI arms race today.
Building an abundant and competitive energy future doesn’t require us to buy into all of Silicon Valley’s promises, but it does require us to start moving faster in case Beijing starts to believe the hype.
Harry Krejsa is director of studies at the Carnegie Mellon Institute for Strategy & Technology. Chris Barnard is president of the American Conservation Coalition.