According to Bank of America analysis, the surging power demands of artificial intelligence data centers threaten to outstrip the generation capacity utilities are planning to add over the next six years. The bank's research indicates that by 2030, data center load growth could exceed available utility capacity by more than 100 gigawatts, creating a significant gap between energy supply and demand in the nation's power infrastructure.
The capacity shortfall is likely to drive data center operators toward alternative solutions, with increased adoption of on-site gas generation and battery storage systems. This shift could reshape the power industry's investment priorities and force utilities to accelerate generation expansion plans beyond current projections to meet explosive computational demand.
The findings underscore how rapidly evolving technology sectors can stress existing infrastructure frameworks and highlight a critical planning challenge for utilities and policymakers. As artificial intelligence expands across enterprises and cloud providers, the mismatch between available grid capacity and actual power consumption will likely intensify pressure for faster deployment of new generation assets and grid modernization efforts.