AI Energy Needs Drive 66% Cost Spike For Natural Gas Plants

The rapid expansion of artificial intelligence and the massive data centers required to support it are driving a historic spike in the construction costs of natural gas power plants. According to recent industry data, the capital expenditure needed to build these facilities has surged by 66% as tech giants like Microsoft and Meta scramble to secure consistent energy sources. This shift marks a significant pivot for a sector that has long prioritized decarbonization through wind and solar power.

The surge in costs is attributed to a combination of supply chain bottlenecks and a "land grab" for specialized equipment. As tech companies compete to bring online the massive amounts of power needed for training advanced AI models, they are finding that renewable energy alone cannot yet provide the 24/7 reliability required for around-the-clock operations. Consequently, the reliance on natural gas is increasing, even as companies face pressure to meet their ambitious net-zero targets.

This trend highlights a growing tension between the tech industry’s environmental goals and its insatiable thirst for computing power. While natural gas burns cleaner than coal, it remains a major source of carbon emissions and methane leaks. Industry watchers are now focusing on how these rising costs will impact the energy market at large and whether tech companies will begin investing in more experimental firm power alternatives, such as small modular nuclear reactors or enhanced geothermal systems, to bypass the volatility of the gas market.

This report was originally published by TechCrunch.