IEA: AI's Energy Appetite Puts Sustainability in Focus
AI is transforming how the world consumes energy, prompting urgent questions about sustainability, power infrastructure and long-term environmental impact.
The International Energy Agency (IEA) has released a report, 'Energy and AI', that examines how AI’s rise is reshaping the global energy landscape and what governments, businesses and energy operators must do to manage this shift responsibly.
The report is the first detailed international analysis of how AI interacts with energy systems, both as a new source of rising electricity demand and as a tool to accelerate sustainable solutions.
Dr Fatih Birol, Executive Director of the IEA, explains: “In recent years, AI has soared to the top of the political and business agenda.
“Once a mostly academic pursuit, it has evolved into an industry with trillions of dollars at stake. Despite significant uncertainties, it is now very clear: AI is coming. In many sectors, it is already here.
“This has major consequences for the global energy sector. There is no AI without energy – specifically electricity. At the same time, AI can potentially transform the sector’s future. However, policymakers and the market have often lacked the tools to fully understand these wide-ranging impacts.
“Recognising this gap, the International Energy Agency (IEA) stepped up to address it by leveraging our expertise in data collection and analysis, as well as our convening power, to inform and strengthen the global dialogue on these issues.”
The growing energy cost of AI
One of the most immediate effects of the AI boom is the steep increase in electricity consumption.
The IEA estimates that global data centres used around 415 terawatt-hours (TWh) of electricity in 2024, or roughly 1.5% of global power usage.
This figure is expected to more than double to 945 TWh by 2030 and could reach as high as 1,720 TWh by 2035, depending on the adoption rate of AI and trends in energy efficiency.
To put that in perspective, the upper estimate would exceed Japan’s current total electricity consumption.
The United States is set to see the largest increase, with data centre electricity use potentially surpassing the consumption of its heavy industry sector.
This spike is driven largely by “digital workloads” such as AI model training, video streaming and cloud computing.
Accelerated servers with specialist chips for AI now operate four times faster than conventional servers. While that improves processing power, it adds to the electricity load exponentially.
The IEA warns this trend is not only unsustainable but risks undermining global climate goals.
However, it also outlines strategies to reduce the environmental impact, including:
Locational flexibility: Building data centres in regions with lower carbon electricity or cooler climates for passive cooling.
Renewable energy integration: Powering AI infrastructure with solar, wind and hydroelectric sources.
Operational flexibility: Adjusting computing loads during off-peak times or when renewable energy is most available.
Without changes to how and where AI infrastructure is powered, sustainability targets will be harder to reach.
AI as a sustainability enabler
Despite its energy demands, AI also offers powerful solutions to help meet climate goals.
The IEA report highlights several ways in which AI can support sustainability and increase energy efficiency across industries.
In the energy and mining sectors, AI improves the precision of exploration, streamlines production and enables real-time environmental monitoring. Automated systems can detect methane leaks, assess resources more accurately and reduce energy waste.
Electricity grids also benefit. AI-based systems can forecast renewable energy generation more accurately, enabling better integration of wind and solar into power systems.
Fault detection systems using AI can cut grid outages by 30% to 50%, increasing reliability and reducing downtime.
The IEA estimates that wider AI adoption in manufacturing alone could result in annual energy savings equivalent to Mexico’s entire energy use today.
In buildings, AI-controlled systems that optimise heating, ventilation and cooling (HVAC) could reduce global electricity demand by 300 TWh each year.
In transport, AI enables:
More efficient route planning
Predictive maintenance of vehicles
Autonomous vehicle systems that reduce unnecessary fuel use
Meanwhile, AI-led cybersecurity can respond to energy infrastructure threats far faster than traditional systems.
AI sensors and satellite data can act up to 500 times faster in threat detection and response.
These innovations position AI as both a challenge and a tool.
The IEA believes that by 2035, AI technologies could cut energy-related emissions by up to 5%.
Building sustainable AI infrastructure through collaboration
To support sustainable AI growth, the IEA is calling for greater cooperation between governments, the tech industry and energy providers.
The report warns that fragmented decision-making and slow policy responses risk allowing AI’s energy use to spiral unchecked.
According to the IEA, the solution lies in three pillars:
Energy mix strategy: Countries must develop a balanced power generation portfolio that supports data centres’ uninterrupted electricity needs. This could include established options like solar, wind and natural gas, as well as newer solutions such as advanced geothermal systems and small modular nuclear reactors (SMRs).
Infrastructure investment: Without expanding and modernising electricity grids, efforts to meet AI’s energy needs will fall short. Smarter, more flexible grids are key.
Cross-sector dialogue: Bridging gaps between energy policymakers, AI developers and infrastructure planners is essential to align on goals and avoid siloed approaches.
“To deliver the energy for AI, countries must also think about their infrastructure,” Fatih says.
“Making this a reality will hinge on the final pillar: bolstering dialogue between policy makers, the tech sector and energy industry.”
The IEA advocates for streamlined permitting and faster connection processes for renewable projects.
It also recommends enhanced public-private partnerships to strengthen digital skills and develop a future-ready energy workforce.
Universities, technology firms and governments should jointly invest in training and research to reduce the current shortage in AI expertise.
Finally, it calls on governments to fund AI solutions that advance low-carbon technologies and to implement frameworks that encourage sustainable AI adoption.
These steps, the IEA argues, will not only control AI’s growing electricity needs but also ensure the technology contributes meaningfully to energy efficiency, affordability and security.
The message is clear: AI’s future must be powered sustainably.
The decisions made now will determine whether AI becomes a force for environmental progress or a challenge to be managed.
Explore the latest edition of Sustainability Magazine and be part of the conversation at our global conference series, Sustainability LIVE.
Discover all our upcoming events and secure your tickets today.
Sustainability Magazine is a BizClik brand
