A new forecast warns gas prices could triple in parts of the US, threatening the economics behind hyperscalers' power strategy.
Hyperscalers have leaned on natural gas to power the buildout of AI data centers, betting on cheap, reliable supply. A new forecast suggests gas prices could triple in some US regions, which would significantly raise operating costs for facilities already running at massive scale.
The exposure comes at a moment when AI infrastructure spending is already under investor scrutiny, adding another variable to the cost equation for companies racing to add compute capacity.
Energy cost volatility is now a direct line item risk for AI economics, not a background utility concern — operators locked into long-term gas contracts should stress-test margins against a tripled-price scenario. Boards evaluating data center capex should demand energy hedging plans alongside GPU procurement plans.
The daily signal, curated. Get it in your inbox.
Subscribe on LinkedIn →