> ## Content Index
> Fetch the complete content index at: https://datacatalyst.org/llms.txt
> Use this file to discover other available public pages before exploring further.

# America Needs Far More Power, and the AI Boom Can Help Pay for It
- URL: https://datacatalyst.org/america-needs-far-more-power-and-the-ai-boom-can-help-pay-for-it/
- Published: 2026-10-20T17:39:04.000Z
- Updated: 2026-10-20T17:39:04.000Z
- Description: Electricity demand is rising faster than at any time this century. Long-term contracts from AI companies can pay for new power plants without pushing those costs onto household bills.
- Author: DCI
- Tags: Energy, #hero-immersive, AI, Data Centers, electricity prices, power grid, US-China competition, state policy

America’s economic growth now depends on building far more electricity, because demand from data centers, factories, homes, and electric vehicles is [rising faster](https://www.eia.gov/pressroom/releases/press582.php?ref=datacatalyst.org) than at any time this century. Utility companies expect peak demand to [rise about 20 percent by 2030](https://gridstrategiesllc.com/wp-content/uploads/Grid-Strategies-National-Load-Growth-Report-2025.pdf?ref=datacatalyst.org), and data centers, many of them built for AI, account for about half of that growth. Three-quarters of large energy users [surveyed by PwC](https://www.pwc.com/us/en/industries/energy-utilities-resources/large-energy-user-survey.html?ref=datacatalyst.org) already say energy constraints are holding back their growth.

This is the moment for America to build far more power from every reliable source, with the AI industry and other large customers driving new demand paying for much of it. The companies building AI need dependable electricity for decades to come and will sign long-term contracts to get it, which is exactly what power developers need to finance new plants. That demand can pay for new supply the whole economy shares, while private investors carry the risk and household costs stay protected.

Whichever country builds power fastest now will lead the world in AI and manufacturing for years to come, and last year [China](https://www.mercomindia.com/strong-solar-buildout-pushes-chinas-power-capacity-to-3890-gw?ref=datacatalyst.org) [added](https://www.eia.gov/todayinenergy/detail.php?id=67205&ref=datacatalyst.org) about ten times as much generating capacity as the United States. However, American companies have the money to compete, and Alphabet, Amazon, Meta, and Microsoft alone [plan to spend](https://www.datacenterknowledge.com/hyperscalers/big-tech-to-spend-650b-this-year-as-ai-race-intensifies?ref=datacatalyst.org) more than three times as much on new facilities and equipment in 2026 as America’s major electric utilities typically [invest annually](https://www.eei.org/News/news/All/electric-companies-to-invest-nearly-$208b-in-2025-to-strengthen-grid-and-drive-economic-growth?ref=datacatalyst.org). Yet as venture investor Chamath Palihapitiya [recently put it](https://podcasts.apple.com/us/podcast/coreweave-crash-ai-trade-cracks-anthropic-ipo-red-flags/id1502871393?i=1000796753063&ref=datacatalyst.org) bluntly, “I think power will be the moat.” Companies can raise capital and buy AI chips, but their ability to put those investments to work ultimately depends on securing enough electricity to run them.

Indeed, long-term contracts from tech companies are the fastest way to finance new power plants, as [Google’s deal](https://www.aep.com/news/stories/view/12314/?ref=datacatalyst.org) for a new fuel-cell plant in Wyoming and [Microsoft’s deal](https://www.utilitydive.com/news/constellation-three-mile-island-nuclear-power-plant-microsoft-data-center-ppa/727652/?ref=datacatalyst.org) to [restart a reactor](https://www.latitudemedia.com/news/restarts-and-uprates-are-powering-the-nuclear-renaissance/?ref=datacatalyst.org) at Three Mile Island in Pennsylvania show. A 20-year contract tells investors, policymakers, and the public exactly who will buy the power, and for how long. Regulators reviewing any large energy project should ask one question: how much new power does it bring, compared with how much it takes?

To protect households and businesses from paying for power plants and lines that nobody uses, data center developers that reserve grid capacity should have to pay for it even if they cancel. In fact, [Ohio already requires](https://www.aepohio.com/company/about/rates/data-center-tariff/?ref=datacatalyst.org) large data centers to commit to paying for most of the capacity they reserve, back that promise with financial guarantees, and pay a steep price to cancel. 

Even with money and customers lined up, shortages of expensive, complex gas turbines and grid transformers can hold up new power plants for years, so America needs every reliable source it can bring online quickly, from gas and nuclear to solar and batteries. For example, GE Vernova’s gas turbine order book is largely [sold out years in advance](https://www.turbomachinerymag.com/view/ge-vernova-gas-turbine-backlog-hits-116-gw-as-power-orders-more-than-double?ref=datacatalyst.org), and large transformers take about [two and a half years to arrive](https://www.foley.com/insights/publications/2026/09/the-grids-bottleneck-isnt-just-money-its-materials/?ref=datacatalyst.org). At the same time, America needs to bring new energy technologies to market, from [advanced geothermal](https://www.eia.gov/todayinenergy/detail.php?id=67204&ref=datacatalyst.org) and next-generation nuclear plants to better energy storage and, eventually, [fusion](https://www.energy.gov/articles/energy-department-releases-finalized-fusion-science-and-technology-roadmap-accelerate?ref=datacatalyst.org). The AI boom could provide both the demand and the financing to help promising innovations reach commercial scale. Meta has agreed to help [fund two new TerraPower nuclear units](https://finance.yahoo.com/news/big-tech-puts-financial-heft-103254759.html?ref=datacatalyst.org), for example, and Google has [agreed to buy power](https://cfs.energy/news-and-media/google-and-commonwealth-fusion-systems-sign-strategic-partnership/?ref=datacatalyst.org) from Commonwealth Fusion Systems’ first commercial fusion plant in Virginia. The right mix will differ by region, so the state utility commissions and local permitting offices that approve most new power plants should let buyers and developers choose whatever gets power online fastest and cheapest.

Right now, the biggest threat to America’s power build-out is a public and political backlash against data centers, and partisan fighting over electricity bills could cost the country its lead. Earlier in 2026, [local opposition](https://www.fortune.com/2026/06/16/data-center-opposition-construction-delays-blocks-report/?ref=datacatalyst.org) delayed or canceled at least 75 data center projects worth more than $130 billion, and the number of opposition groups doubled to more than 800 across 49 states. In August 2026, three-quarters of American voters said they would [oppose a new data center near their home](https://heatmap.news/daily/data-center-opposition-poll-collapse?ref=datacatalyst.org), up from about four in ten a year earlier. Campaigns and outside groups in both parties [have spent more than $45 million](https://www.opb.org/article/2026/09/05/data-centers-have-republicans-and-democrats-on-the-same-page/?ref=datacatalyst.org) on midterm ads that mention data centers, many blaming them for rising electricity bills.

Ironically, blocking projects keeps in place the very shortage that is driving up people’s bills.

Voters will accept new data centers and power plants when developers pay their own way and utilities show the effect on household bills, which are [already up about 5 percent this year](https://www.utilitydive.com/news/data-centers-electricity-residential-rates-eia/820342/?ref=datacatalyst.org). President Trump’s [Ratepayer Protection Pledge](https://www.abc15.com/science-and-tech/artificial-intelligence/trump-administration-touts-more-support-for-reducing-datacenter-power-costs?ref=datacatalyst.org), signed by Google, Amazon, Meta, and over 20 governors, asks tech companies to do exactly that by bringing their own power and paying for grid upgrades. Utilities can help close the deal by publishing what every new project means for a typical household’s bill.

The Trump administration has made faster grid connections for data centers and factories a national priority, and [federal regulators have ordered](https://www.foley.com/insights/publications/2026/06/ferc-issues-six-tailored-show-cause-orders-to-accelerate-large-load-interconnection/?ref=datacatalyst.org) the grid operators serving 200 million Americans to speed them up. Today, connecting a large new customer such as a data center typically [takes at least five years](https://www.americanactionforum.org/insight/ferc-data-center-orders-accelerate-grid-connection/?ref=datacatalyst.org). In June 2026, the Federal Energy Regulatory Commission (FERC) [gave](https://www.ferc.gov/news-events/news/ferc-launches-aggressive-targeted-action-speed-large-load-integration?ref=datacatalyst.org) six regional grid operators 60 days to justify their rules or change them. The commission also [told the operators](https://www.ferc.gov/news-events/news/ferc-launches-aggressive-targeted-action-speed-large-load-integration?ref=datacatalyst.org) to make it easier to build power plants right next to the data centers and factories they serve. But, ultimately, states still decide whether and where individual plants are built and what retail customers pay.

The decisive fight over America’s power supply is in the states, which are splitting into builders like Texas and blockers like New York. [Texas passed a law](https://www.datacenterdynamics.com/en/news/texas-governor-approves-bill-mandating-data-centers-to-contribute-to-costs-of-interconnection/?ref=datacatalyst.org) in 2025 that requires large data centers to pay part of their grid connection costs. In contrast, [New York Governor Kathy Hochul ordered](https://broadbandbreakfast.com/new-york-becomes-first-state-to-impose-data-center-moratorium/?ref=datacatalyst.org) a one-year pause on permits for large data centers in mid-2026, the first statewide moratorium in the country. Lawmakers in at least a dozen other states [have proposed similar bans](https://www.stateside.com/blog/rise-data-center-moratoria?ref=datacatalyst.org). The Texas approach adds power and protects households. The New York approach adds no new supply and leaves the shortage for someone else to solve. President Trump called New York’s ban “[a terrible decision](https://www.wshu.org/2026-07-16/hochul-pushes-back-as-trump-blasts-her-data-center-moratorium?ref=datacatalyst.org).”

Business leaders, policymakers, and families can judge every state’s energy policy with three measures: how fast new customers connect, how much reliable power gets added, and what happens to electricity bills. Every state should publish all three each year, so companies, voters, and lawmakers can make decisions about power based on facts, rather than political spin and negative ads.

In the end, electricity will decide whether America or China leads the world in AI and advanced manufacturing, and which American communities win the jobs and investment that come with that lead. America has the capital, the customers, and the technology to build the power it needs. What the country needs now is leaders in both parties with the will to build.