The Rising Cost of Electricity: A Direct Consequence of Data Centers and AI Demand
“I could not keep paying such an invoice,” says a woman from Baltimore. In regions of the United States close to high concentrations of data centers, the wholesale price of electricity has soared by up to 267% in just five years. Surprisingly, consumers are bearing these escalating costs in their utility bills.
Short. An exhaustive Bloomberg analysis has correlated the prices across 25,000 points of the American electricity grid with the proximity of the country’s data centers. The result is alarming: more than 70% of nodes that recorded significant price increases are situated less than 80 kilometers from these data hubs.
AI and the Price of Light. According to the International Energy Agency , electricity consumption by data centers is set to double to 945 TWh by 2030 , surpassing the energy demands of entire countries like France or Japan . This surge in demand is primarily driven by advancements in artificial intelligence (AI) .
While major technology companies argue that their data centers provide investment, quality jobs, and tax revenues that benefit local communities, the residents living near these energy-intensive facilities have been grappling with less-publicized consequences.
A Matter of Supply and Demand. The energy demands of AI-driven data centers are so substantial that they have pushed electric capacity auctions to record highs. This trend is particularly evident with the largest network operator in the U.S., PJM Interconnection , which spans 13 states. In its latest annual auction, data center demand led to sales exceeding $7.3 billion , marking an 82% increase .
The reasoning is straightforward: energy demand is outpacing new generation capabilities. However, the implications extend beyond merely invigorating the installation of new energy sources. The wholesale prices paid for electricity to secure future supply directly affect both corporate and household utility bills.
David Against Goliath. To grasp the local impact, one can travel to Virginia , home to the largest concentration of data centers worldwide. Estimates suggest that between 30% and 40% of global internet traffic passes through servers in this region. The Dominion Energy company anticipates a staggering 85% increase in demand by 2039 due to these facilities.
If it weren’t for the data centers, demand growth would only be 10% . However, as it stands, new infrastructure—such as high-voltage towers and substations—will be necessary, and the costs will be distributed among all customers of the electric company, impacting everyone from small apartment residents to giant data centers.
This Is Not Just an American Issue. Although the epicenter of this crisis lies in the U.S. due to its dominating AI sector, this problem is emerging globally. Technology firms are keen to expand their infrastructure worldwide, creating similar pressures on electrical grids everywhere.
In the United Kingdom , the burgeoning demand from data centers is projected to increase electricity prices by 9% by 2040. Similarly, in Malaysia , the government has had to raise electricity rates for data centers as new facilities challenge the nation’s supply capabilities.
The Social Cost of Technological Progress. Data centers require construction and maintenance, which generates employment. In Virginia alone, they contribute a staggering $9.1 billion to the state’s GDP. Companies like Microsoft, Amazon, and Google are reinforcing their investments in renewable energy and more efficient technology, helping to alleviate some of the costs associated with their energy consumption.
Nevertheless, as long as the electricity grid struggles to keep pace and utility bills continue to rise, the debate surrounding the societal cost of AI innovation will remain a burning issue, providing fertile ground for discussion regarding equitable energy distribution and responsible technological advancement.
Image | Rawpixel
For further insights, visit Xataka, where Google has finally disclosed its energy and water consumption rates for AI operations, revealing the complexities behind modern technological advancements.