AI Data Centers and Their Impact on the Electric Grid
What: AI Data Centers and Their Impact on the Electrical Grid
Where: Faculty of Electrical Engineering, University of Ljubljana
When: 23th of September, 2025
On September 23, the Faculty of Electrical Engineering hosted a lecture by Dr. Edvina Uzunovic from Worcester Polytechnic Institute, organized under the IEEE Power & Energy Society Slovenian Chapter and Laboratory of energy policy. The lecture addressed one of the most pressing topics in today’s energy and digital transformation: the rising impact of AI data centers on electricity consumption and grid stability.
AI and the New Energy Challenge
Generative AI and big data are accelerating the growth of data centers worldwide. While enabling innovation, this expansion brings unprecedented energy demands. As Dr. Uzunovic explained, a single AI query can consume nearly 10× more electricity than a standard internet search, and 40–50% of data center energy is spent on cooling alone.
In the United States, data centers already use about 4.4% of national electricity (2023). By 2028, projections suggest this share could rise to 6.7–12%, creating significant carbon and water footprints. Importantly, modular data centers can be built within 1–3 years, while new renewable energy projects often require 3–7 years, a timing gap that complicates energy planning.
Challenges for the Grid
The rapid and often secretive development of AI data centers creates uncertainty for grid operators and regulators. Key challenges highlighted in the lecture included:
-
Grid stability and capacity constraints
-
Uncertain load forecasts due to fast deployment
-
Regulatory hurdles around integration and siting
-
Ensuring resilience and ride-through capabilities during disturbances
To meet these challenges, Dr. Uzunovic outlined several trends and potential solutions:
-
Expanding renewables (solar, wind, hydro, geothermal)
-
Using natural gas as a transitional fuel
-
Exploring hydrogen and fuel cells as backup sources
-
Renewed focus on nuclear energy, including small modular reactors (SMRs), as a stable, carbon-free option to support both climate goals and the digital revolution
Global and European Context
While the lecture focused on the U.S. energy mix and deployment timelines, the discussion also addressed Europe’s fast-growing data center landscape, raising similar questions about how to balance digital growth with sustainability and grid resilience.
Dr. Uzunovic concluded with a clear message: energy production must be decoupled from environmental impact, and clean energy must become abundant and affordable if society is to meet the challenges of AI-driven digitalization.
