Hyperscalers face price volatility in pivot to natural gas for AI

While current natural gas prices hover between $2 and $4.50 per million BTUs, Noreva projects costs could triple to exceed $10 in specific U.S. hubs. The firm’s CEO, Peter Gardett, argues that the market is currently lulled into a false sense of security, ignoring the mathematical reality of a narrowing supply-demand gap. By building their own gigawatt-scale generation facilities, hyperscalers are straying from their traditional asset-light models and exposing themselves to commodity market volatility they are ill-equipped to manage.

The push into fossil fuels has been driven by the availability of cheap, stranded gas in regions like West Texas. However, as new pipeline infrastructure connects these localized markets to international export terminals, the era of discounted fuel is ending. Because fuel typically accounts for half of the operating cost of a power plant, these price swings threaten to inflate the cost of AI tokens or force companies to rely more heavily on increasingly strained public grids.

Beyond the financial risk, this strategy risks intensifying public scrutiny. With 80% of consumers already anxious about the impact of data centers on their utility bills, a rise in natural gas prices could broaden the backlash against big tech’s infrastructure expansion. Investors are already noticing the shift, with analysts expecting future earnings calls to begin addressing the direct correlation between energy market fluctuations and corporate bottom lines.