According to Fortune, research indicates that the rapid expansion of AI data centers in the U.S. has significantly increased public electricity costs
According to research cited by Fortune, the rapid expansion of AI data centers in the U.S. has led to a substantial increase in public electricity costs.
The PJM market monitoring agency estimates that the additional electricity demand from data centers will impose approximately $23 billion in extra costs on electricity users, with this impact expected to last at least until the end of 2028.
Although tech companies have pledged to bear some infrastructure costs, expenses for public facilities such as transmission lines, substations, and grid upgrades are typically shared uniformly, with some burden still passed on to residents and ordinary commercial users. Some data centers have reduced peak load sharing through flexible load adjustments, but overall electricity consumption remains high, and their actual cost burden is lower than the pressure on the grid.
In market mechanisms, AI-driven demand stimulates data center operators to expand, putting pressure on electric utilities and regulators to recover costs, with funds being indirectly transferred from residential users to large tech infrastructures. In response to these events, energy regulation needs to reshape pricing mechanisms to balance growth with fair burden sharing.
Source: Public Information
ABAB AI Insight
PJM, as the largest regional grid operator in the U.S., has long managed the power supply and demand in the Mid-Atlantic and Midwest regions. It has raised capacity auction prices multiple times in the past two years due to the predicted surge in data center loads, and this $23 billion estimate continues its monitoring of the impact of AI demand.
In terms of capital flow, tech giants are mobilizing huge investments through data center construction, but the costs of upgrading the public grid are distributed among all users through regulation. The motivation is to rapidly deploy computing power to gain a leading position in AI while avoiding the full burden of infrastructure spending.
Similar historical cases, such as the early expansion of cloud computing raising regional electricity prices, indicate that the current U.S. energy system is in the early stages of an explosion in AI physical infrastructure, with data centers becoming the core driver of new load additions.
Essentially, this represents capital concentration, where AI giants achieve low-cost expansion through shared public grids, leading to a shift in pricing power from regulators and utilities to high-energy-consuming tech platforms. The uniform sharing model amplifies externalities, forcing regulators to redefine the cost responsibilities of large loads.
ABAB News · Cognitive Law
- The external costs of computing power expansion are ultimately borne by society as a whole.
- Flexible loads reduce peaks but cannot mask the overall system burden of total consumption.
- The greater the infrastructure benefits, the more the sharing disputes become central to regulation.