The International Monetary Fund (IMF) has expressed concerns that the surging demand for electricity to power Artificial intelligence (AI) data centers could create new challenges for energy infrastructure, prices, and emissions globally.
The IMF, in a statement published on Tuesday, noted that AI technologies are reshaping productivity, employment, and investment patterns worldwide, with the potential to significantly raise the average pace of annual global economic growth.
However, this economic potential is closely tied to a steep rise in electricity demand, particularly from data centers that power AI models and cloud computing.
AI’s power appetite is growing rapidly
According to recent estimates by the Organization of the Petroleum Exporting Countries (OPEC), data centers consumed about 500 terawatt-hours (TWh) of electricity in 2023, a figure that more than doubled the annual average from 2015 to 2019.
- IMF said That number could triple to 1,500 TWh by 2030, rivaling the current electricity consumption of India, the world’s third-largest electricity consumer.
- To put that in perspective, data centers already use as much electricity as France or Germany, and by 2030, could consume 1.5 times more power than electric vehicles (EVs).
- The United States, which hosts the world’s largest concentration of data centers, is expected to lead the growth in energy usage.
- According to projections by McKinsey & Co., electricity consumption from US server farms could exceed 600 TWh by 2030, more than triple current levels.
- The continued boom in AI-driven applications and the expansion of cloud infrastructure are driving this energy surge, raising urgent questions for policymakers about how to prepare national grids for the future.
Economic growth at risk if energy supply lags
The IMF warns that insufficient investment in electricity infrastructure could threaten the very benefits AI promises.
While increased demand could incentivize more electricity production and only modestly raise prices, sluggish responses might result in steep energy costs that hurt consumers, businesses, and ultimately stall AI-related growth.
The global economy could also face delayed or reduced investment in AI if electricity becomes unreliable or unaffordable for major tech infrastructure operators.
“Increasing electricity demand from the technology sector will stimulate overall supply, which, if responsive enough, will lead to only a small increase in power prices.
“More sluggish supply responses, however, will spur much steeper cost increases that hurt consumers and businesses and possibly curb growth of the AI industry itself,” the IMF stated.
Climate costs
Beyond the economic risks, the environmental implications are stark. The IMF projects that AI-related electricity demand could generate an additional 1.7 gigatons of global greenhouse gas emissions between 2025 and 2030, roughly equivalent to Italy’s five-year energy-related emissions.
This highlights the dual challenge facing governments: unlocking AI’s potential while managing its environmental footprint.
- The IMF observed that efficient, open-source AI models like DeepSeek are helping reduce computing costs and energy usage per task.
- However, it noted that lower costs often drive higher adoption and usage, which can neutralize these energy savings.
- Additionally, more advanced reasoning models tend to be significantly more energy-intensive, further complicating the equation.
The way forward
With the full impact of AI on electricity demand still uncertain, the IMF calls for proactive policy measures.
“Implementing policies that incentivize multiple energy sources can enhance electricity supply, help mitigate price surges and contain emissions,” it stated.
The IMF added that governments and private sector players must collaborate to scale up electricity supply through diversified sources, including renewables, to cushion price shocks and contain emissions.