Artificial intelligence (AI) is emerging as a key driver of productivity and economic growth, reshaping employment and investment patterns globally. According to scenarios outlined in the International Monetary Fund’s (IMF) April 2025 World Economic Outlook report, AI has the potential to boost the average annual pace of global economic growth.
However, AI systems require significant amounts of electricity to power data centers, which could place increasing pressure on global energy grids. This growing demand has major implications for worldwide electricity consumption.
According to recent estimates from the Organization of the Petroleum Exporting Countries (OPEC), global data centers consumed over 500 terawatt-hours (TWh) of electricity in 2023. This represents more than double the consumption level between 2015 and 2019, and is projected to triple to reach 1,500 TWh by 2030, according to OPEC projections.
As illustrated in this week’s chart, data center electricity consumption has already reached levels comparable to the annual electricity usage of Germany or France. By 2030, this consumption could match India’s total electricity demand, which ranks third globally. These figures would significantly surpass electricity consumption attributed to electric vehicles (EVs), with data centers using 1.5 times more electricity than EVs by the end of this decade.
The fastest-growing region for data center energy consumption is the United States, which hosts the highest concentration of data centers worldwide. According to McKinsey & Co.’s medium-demand scenario projection, electricity needs for server farms in the U.S. are expected to more than triple, exceeding 600 TWh by 2030.
The rapid expansion of cloud data storage facilities responding to AI queries highlights the urgency for policymakers to develop effective energy strategies that can meet rising demand. The growing energy requirements of the technology sector will stimulate overall supply, potentially leading to modest increases in electricity prices if supply responds adequately. However, slower supply responses could result in significant price hikes that negatively impact consumers and businesses, potentially slowing growth in the AI sector itself.
If current energy policies remain unchanged, AI-related electricity demand could add approximately 1.7 gigatons of global greenhouse gas emissions between 2025 and 2030 – equivalent to Italy’s total energy-related emissions over five years.
The demand for computing power and electricity from AI platforms remains uncertain. While efficient open-source AI models like DeepSeek reduce computing costs and electricity demand, lower costs also increase AI adoption. Energy-intensive reasoning models, meanwhile, drive up electricity consumption. The net effect on electricity demand remains unclear, potentially delaying energy investments and causing price increases.
Governments and businesses must collaborate to ensure AI reaches its full potential while minimizing costs. Implementing policies that encourage the use of diverse energy sources can enhance electricity supply, help mitigate price increases, and reduce emissions.