Artificial intelligence is reshaping the global energy equation, driving unprecedented electricity demand while unlocking new efficiencies across the value chain. This article explores how governments and industry are racing to secure scalable, lower-carbon power through LNG and hydrogen, while embedding AI to optimise operations and accelerate the energy transition.
The rise of artificial intelligence (AI) and its massive energy needs is presenting governments and industries alike with supply challenges and opportunities that are reshaping the entire energy ecosystem and geopolitical landscape.
AI-driven data centres consumed approximately 1.5% of global electricity in 2024, with projections suggesting this could double by 2030, the International Energy Agency (IEA) estimates. At the same time, the energy sector is witnessing an unprecedented rollout of AI-based technologies and solutions that are enabling improved operations, efficiency and more seamless integration and supply.
At the asset level, studies show that digital tools powered by AI are enabling energy companies to lower operating costs by up to 25%, improve productivity by 8%, and increase energy efficiency by as much as 8%, according to the IEA. These are not incremental shifts; they are foundational gains that are fundamentally altering how energy is produced and consumed.
of global electricity consumed by data centres in 2024
Energy companies’ lowered operating costs through AI-powered digital tools
“AI is in the minds of everyone. Long gone are the days when AI was a science experiment. The way that we are now across these industries and going forward, AI is truly the reliability engine behind any operations. All sectors are actually having that in mind. The question then becomes how do we make that with the good data, bringing people along that can trust the reliable responses that you get out of it,” said Fabricio Sousa, President of Worley Consulting Global, while speaking at Gastech 2025.
Enabling AI electrification for national security
Given the increasing impact and even greater unmet potential of AI on the global energy system, it comes as no surprise that AI is being viewed by government and industry leaders alike as a major opportunity to advance their existential goals.
For the United States, leading on the development and integration of AI is a question of national security, as Doug Burgum, US Secretary of the Interior, recently revealed. Speaking at Gastech, he said, “So this is the race to win the AI arms race. It's really a race to see who can add the most electricity. Because again, we're not energy transitioning. This is about energy addition.”
In order for the US to secure its AI leadership, it is prioritising increasing its energy supply as quickly as possible in order to power the data centres that AI models run on.
“Electricity now is the new playing field because electricity, for the first time, can be turned into intelligence, converted to intelligence, and when you have that prospect, there's a little bit of who has the most electricity is going to win,” Secretary Burgum added.
“Electricity now is the new playing field because electricity, for the first time, can be turned into intelligence, converted to intelligence, and when you have that prospect, there's a little bit of who has the most electricity is going to win,”
- Doug Burgum, US Secretary of the Interior
In July, the US government announced the ‘America’s AI Action Plan’, which outlines its objective of achieving global technological dominance in AI. Other countries with national plans for AI leadership include China, the United Kingdom, France, and the United Arab Emirates, whose National Strategy for Artificial Intelligence 2031 aims to make AI a cornerstone of the country’s development and transform the country into a global hub for responsible and innovative AI.
Rapid and scalable power for AI through LNG
Providing that electricity at the rate and speed required by the world’s burgeoning data centres requires leveraging all available sources of energy. One of the most promising types of energy that can be brought online quickly and affordably is liquefied natural gas (LNG).
As data centres require continuous, high-capacity electricity to power their general processing units and 24-hour operations, LNG is being positioned as a critical part of the new energy mix. Compared to renewable energies like solar and wind, LNG is relatively more stable, scalable, and dispatchable. It is also lower-carbon than coal and oil, making it a ‘bridge fuel’ that can support more sustainable energy consumption, particularly when paired with carbon-capture technologies.
“AI, like we see it in all parts of human life today, is changing and altering that and disrupting it in many respects. With the glut of energy that is required, it's not only the glut, it's the speed at which energy is required to support the powering and the expansion of data centres…The interplay between AI and LNG has been at an all-time intersection, primarily because of the power consumption that is associated with AI data centres. And I don't believe right now, even though we have a large installed base globally of renewables, it's certainly not the dispatchable type of power that's going to meet the requirements of data centre capacity,” said Devan Pillay, Cluster President, Schneider Electric.
Hydrogen for decarbonised remote AI power solutions
For more carbon-conscious and remote deployments, experts are also eyeing hydrogen as one of the many sources of energy that will help meet the growing energy demand of AI, alongside the general increased energy consumption.
Hydrogen supply can be low- to zero-carbon, making it an energy source that aligns with the decarbonisation requirements of many countries and industries. It is also compatible with off-grid energy solutions, enabling supply to remote data centres. Not being reliant on the grid also allows hydrogen-based setups to be deployed sometimes twice as quickly as traditional grid-connected facilities, supporting the rapid expansion of AI.
Ivana Jemelkova, CEO, Hydrogen Council, said: “There's no doubt that the energy requirements of artificial intelligence and this boom in data centres that underpins it, that's a natural opportunity for hydrogen. Not only because it can deliver benefits on the sustainability front, but it's really that flexibility and resilience where hydrogen can truly shine. So we have high hopes for that. We believe that hydrogen is a very, very good fitting solution.”
This compatibility has already resulted in a number of hydrogen-powered data centres being developed, like ECL’s 1 megawatt facility in Mountain View, California. Other companies like GeoPura are working to deploy hydrogen power units to support data centres with access to limited grid capacity.
As data centres multiply and machine learning models grow more complex, the pressure on global energy systems will intensify. Governments and organisations that respond to this challenge with flexibility, urgency, and realism stand to lead this transformation. In this new era, the race is not only to develop smarter algorithms but to power them sustainably and reliably.