The AI boom is creating a fundamental culture clash between fast-moving tech companies and methodical utility providers forced to work together on unprecedented infrastructure demands. Tech giants accustomed to rapid deployment are discovering that power infrastructure planning takes five to ten years, with some utilities projecting seven-year waits for new grid connections.
Recent friction has emerged over financial requirements and timelines. American Electric Power sparred with Amazon, Microsoft, and Google over data hub connections, ultimately reaching an agreement requiring tech firms to pay for at least 85% of projected energy needs regardless of actual usage. Climate goals add another layer of complexity, as tech companies seeking clean power face utilities planning new natural gas plants to meet surging demand.
“Utilities think of things in 10-year cycles, meanwhile our customers think of things in 12-month cycles. These two galaxies are really colliding now.”
— Raul Martynek, CEO of DataBank
DataBank, which recently raised $2 billion for new facilities, exemplifies companies navigating this challenging intersection. The mismatch between tech’s “move fast and break things” mentality and utilities’ deliberate infrastructure approach represents a critical bottleneck in AI development.
See the original analysis on Bloomberg.
