TreasureX Researchq-fin.GNAI's Power Problem: Why Data Centers May Drive Up Electricity Prices
Alexander Crosier, Kyle Onghai, Ronnie Sircar · September 8, 2026
Artificial intelligence is booming, but it has a hidden bottleneck: electricity. Data centers that power AI models consume vast amounts of power, and the grid isn't keeping up. This new research from arXiv models the clash between surging data-center demand and sluggish power supply growth.
The authors build a framework that links data-center load growth, available generation capacity, and market prices. They first use a simple deterministic model to show how different assumptions about demand and supply growth affect electricity prices. Then, they treat both data-center expansion and new power plant construction as random processes, producing probability distributions for supply, demand, and prices instead of a single forecast. Finally, they frame the decision of how much new generation to build as a stochastic control problem: a profit-maximizing investor chooses how aggressively to invest in new capacity.
The key insight: even when rapid demand growth signals a need for more power plants, uncertainties—like load forecasts, construction delays, and the risk that overbuilding will depress future prices—can discourage investment. This could lead to persistently higher electricity prices, which would raise operating costs for data centers and potentially squeeze profit margins for AI companies and their investors.
For retail investors, this means that the AI boom isn't just about chips and software; it's also about energy infrastructure. Companies that can secure reliable, cheap power may have a competitive edge, while those exposed to volatile electricity markets could face headwinds.
Source: arXiv
Educational content only, generated with AI assistance from public research abstracts. Not investment advice. Thank you to arXiv for use of its open access interoperability.