The impact of AI on global energy systems
The increased use of artificial intelligence (AI) has led to an increase in the power demand for data centers. AI can have a much broader impact on energy systems over time, but it affects both energy supply and the overall economy's energy demand
AI's energy demand
Increased power demand from data centers driven primarily by increased use of AI can potentially significantly increase power demand, despite varying degrees across different regions and countries.
in Current trajectoryUsing about a tenth of the data center's electricity demand growth until 2035, the impact will vary widely across regions. For example, increasing demand for data centers accounts for 40% of US electricity demand growth over the next decade (see the Power Sector).
However, such forecasts of data center power demand are highly uncertain. In part, the demand depends on AI evolution and adoption rates. However, it also depends critically on the energy efficiency of data centers, which has been rising dramatically in recent years. Digital data traffic increased more than 25 times between 2010 and 2024, but data center energy usage only doubled over that period.
While some drivers of these past efficiency improvements (e.g., a broad shift from on-premises data centers to the cloud) may not be able to do any further, other innovations, including ongoing advances in chip design and AI programming, could drive improvements in data center energy efficiency over the coming years.
The impact of AI on the energy sector
However, the demand for electricity in data centers is only a narrow aspect of the possibility of AI's impact on the global energy sector. AI can have a major impact on energy supply over time.
AI is already widely used in the oil and gas industry. For example, we are improving and accelerating exploration through better analysis of geological structures. It is also used to plan and design new oil and gas wells, improving the subsequent operation and efficiency of those facilities.
AI could also accelerate low-carbon energy innovation, for example, through the development of new materials for solar panels and carbon capture, the development of new battery chemistry, or the design and improvements in efficiency of low-carbon hydrogen systems. More fundamentally, continuing advances in AI could unlock major technological breakthroughs in low carbon energy supply, including the development of new advanced biofuels and even usable nuclear fusion.
Accelerated use of AI may improve the efficiency in which energy systems operate. It can enable more efficient planning and operation of electrical grids, helping to forecast demand spikes and optimize battery storage deployment. It helps to make the grid “smart” and allows for more efficient aggregation of large quantities of small assets, such as EVs, rooftop solar, and smart thermostats. And it is already being used in both fossil fuel facilities and electricity systems to better predict obstacles and obstacles.
The broader impact of AI on the global economy and the energy system
However, this impact of AI on energy supply is only an energy sector-specific example of the broader potential impact of technology on the global economy, which has a wider and greater impact on the energy system.
AI could effectively improve global economic growth if it speeds up productivity growth. This can occur in a wide range of ways, including automating tasks over an increasing range of tasks, more efficient use of physical assets, accelerated innovation and new discoveries. Currently, there are very broad estimates of the sizes that can be of these effects. One of the most recent OECD surveys1 The reported estimates vary from moderate effects (US GDP levels will only increase by about 1% over the next decade) to an increase of 2.5pp per year to US productivity growth.
Improved AI-driven productivity growth can have a huge impact on energy demand. The average estimated value of the OECD survey – an increase in productivity of approximately 1.2% per year – is assumed to be visible at the global level, increasing total energy demand by around 15% by 2035, and global energy efficiency continues to improve at historical average speeds. It is 20 times the increase in power demand for data centers Current trajectory.
However, it is probably unrealistic to assume that the energy efficiency of the world economy itself is not affected by major advances in AI. AI helps optimize manufacturing processes and accelerate the development of more energy-efficient products, leading to significant improvements in industrial efficiency. Transportation departments can use AI to better manage traffic and optimize routes. It also significantly reduces energy usage to heat and cool the building. Recent IEA Analysis2 It suggests that widespread application of AI to improve energy efficiency across the economy could have a very significant impact on global energy demand.
Assumptions regarding the current orbit and AI impact below 2°
Current trajectory and Less than 2° Both assume a moderate boost to productivity and GDP growth through the use of AI for outlook. Furthermore, the scenario does not explicitly incorporate AI-driven technical breakthroughs that are critical to energy supply. However, given the rapid pace of development in the design and use of AI applications, estimates of these AI effects are currently very uncertain and their impact can be much larger. Uncertainty in the scale of the final size is increasing, focusing solely on the future power needs of data centers.
