Large technology companies are increasingly turning to natural gas to support the power needs of their expanding AI data centers. But a new forecast from energy research firm Noreva suggests that this strategy may become more expensive than expected.
According to the report, natural gas prices in some parts of the United States could rise sharply in the coming years as demand from hyperscalers meets slower supply growth and stronger liquefied natural gas exports. In some hubs, prices could move above $10 per million BTUs, far above today's roughly $2 to $4.50 range.
Companies including Meta, Microsoft, Google, and Amazon have recently announced major gas-powered energy projects tied to data center expansion, especially in Texas and Louisiana. The shift reflects how AI infrastructure is pushing major cloud players deeper into energy planning and long-term fuel risk.
Noreva CEO Peter Gardett says the market may be underestimating how quickly conditions can tighten. He points to the growing link between domestic gas supplies and global export demand, along with rising AI electricity needs, as key forces reshaping pricing.
If gas costs climb, the economics of self-powered AI campuses could change quickly. Higher fuel prices may increase operating expenses and influence where future data centers are built, how they are powered, and how energy markets respond to digital growth. The next phase of AI infrastructure could help redefine the relationship between computing and energy for years to come.
