The tech industry is notorious for its cycles of intense “fear of missing out,” but the current AI boom has triggered a scramble unlike any before. To keep pace with the exponential energy demands of artificial intelligence, industry giants are moving beyond software and into heavy infrastructure, securing massive amounts of natural gas to keep their data centers humming.
A Massive Expansion in the American South
The rush for power is centered in the southern United States, home to vast shale gas reserves. Major players are now commissioning their own power plants to ensure a steady energy supply:
- Microsoft is partnering with Chevron and Engine No. 1 to develop a plant in West Texas capable of producing 5 gigawatts (GW) of electricity.
- Google has confirmed a collaboration with Crusoe to build a 933-megawatt facility in North Texas.
- Meta is significantly expanding its Hyperion data center in Louisiana, adding seven gas plants to reach a staggering 7.46 GW capacity—enough to power the entire state of South Dakota.
The Cost of the Energy Rush
This sudden demand has strained global supply chains. According to Wood Mackenzie, the price of power turbines is expected to jump 195% by the end of this year compared to 2019 levels. New orders are currently backed up until 2028, with delivery times stretching to six years.
The “Behind the Meter” Illusion
By building plants that connect directly to data centers—a practice known as “behind the meter” operations—tech companies claim they are avoiding strain on the public electrical grid. However, they are simply shifting that pressure to the natural gas grid.
The Risks of a Finite Bet
While U.S. gas supplies are currently plentiful, production growth in major shale regions is beginning to slow. This heavy reliance on a finite resource carries significant risks:
- Price Volatility: While tech giants haven’t disclosed their contract terms, they remain vulnerable to market swings that could inadvertently drive up electricity costs for everyone.
- Industrial Competition: Sectors like petrochemicals cannot easily switch to renewables and may find themselves competing with data centers for limited fuel.
- Climate and Reliability: Extreme weather, such as the 2021 Texas freeze, can cause wellheads to fail. In a crisis, suppliers may have to choose between keeping AI models running or heating residential homes.
By betting big on natural gas, the tech industry is proving that the digital world is still very much tethered to physical, finite resources. Whether this gamble pays off depends on whether the AI fever lasts longer than the fuel supply.
