The White House sounded the alarm in July 2025: without $1.4 trillion in new infrastructure investment, electricity prices could surge as much as 58% by 2030. The culprit is straightforward. Demand for power from AI data centers and Bitcoin mining is accelerating faster than the grid can expand to meet it.
According to Axios reporting cited in the warning, power demand is projected to grow up to tenfold between now and 2030. That's a pace the existing electricity system was never designed to handle. The buildout of new generation capacity, transmission lines, and storage takes years. Permitting alone can stretch into decades for nuclear plants. Grid modernization, by contrast, operates on timelines measured in political cycles and investor returns, not technological possibility.
Bitcoin miners and AI operators face the same hard constraint: power is finite and expensive. Bitcoin mining already concentrates in regions with cheap electricity, typically near hydroelectric dams or coal plants with spare capacity. As those margins tighten, miners will hunt for newer, faster sources. AI data centers are doing the same, bidding aggressively for available power and sometimes signing long-term contracts directly with utilities or independent power producers.
The structural tension is real. Neither industry can throttle demand downward without sacrificing growth or profitability. Grid operators can't spin up capacity fast enough to avoid shortages. The gap widens.
Even prominent voices in energy and technology have struggled to articulate a solution that clears the math. New nuclear plants take a decade or more to permit and build. Renewable capacity is growing but remains intermittent without storage at scale. Gas and coal plants face rising permitting friction and pressure to retire.
The White House figure assumes significant new investment arrives. If it doesn't, or arrives slower than projected, the squeeze accelerates. Higher electricity prices ripple through both industries. For Bitcoin miners operating on tight margins, a 58% jump in power costs could force consolidation or migration to regions with more capacity. For AI operators paying billions to build data centers, cost inflation cuts into returns and may shift where new facilities get built.
The gridlock isn't a technical mystery. It's an infrastructure problem disguised as an energy problem. The timeline to resolve it runs well past 2030, and the capital required is enormous. Until new generation and transmission get built and operational, competition for existing capacity will only sharpen.