How the UK’s Plans for AI Could Derail Net Zero
-
Loading... - Arbi Mounir
- 22 Jan 2025
- 45 Views
- 0 Like
- 0 Comment
The UK government’s ambitious goal to increase public-controlled artificial intelligence (AI) computing power twentyfold by 2030 poses significant challenges to the nation’s net zero targets. This increase in computing power is expected to substantially raise electricity demand, raising critical questions about whether the current renewable energy supply can meet this demand while also supporting the electrification of sectors like heating and transport, which must be fully decarbonised by 2050.
AI systems are inherently energy-intensive due to their reliance on vast amounts of computing power. The process of creating and using AI involves training models and algorithms, which requires significant computational resources. Once trained, these AI models must analyze new data, further contributing to their energy consumption. As AI technology advances, the demand for computing power has surged, creating a bottleneck that could hinder further development.
The UK’s national AI strategy, published in 2021, acknowledged the necessity of increasing computing power capacity to unlock AI’s full potential. However, this increase in capacity comes with substantial energy implications.
Data centres, which are essential for storing, processing, and distributing data, are major consumers of electricity. They play a crucial role in every stage of AI development, from training complex models to running predictions. According to the International Energy Agency, data centres account for approximately 1%-1.3% of global electricity consumption. Recent estimates suggest that developing sophisticated AI systems may require a fourfold increase in computing power annually, with the total data needed for AI training rising by 2.5 times each year.
In the UK, AI and related infrastructure consumed around 3.6 terawatt-hours (TWh) of electricity in 2020. If this consumption increases twentyfold, as the government aims, it could reach 72 TWh by 2030. This figure would represent over one-quarter of the UK’s total electricity consumption in 2021, which was approximately 261 TWh.
Moreover, the energy demands of AI extend beyond data centres. Devices utilizing AI—such as smart home sensors, gas and electricity meters, routers, and streaming devices—add to the overall energy consumption, complicating estimates of total demand.
While the UK has made significant progress in renewable energy production, with wind and solar power contributing over 40% of electricity in recent years, projections indicate that global renewable electricity supply may not keep pace with the surging demand driven by digital data growth. Research published in the journal Energy Policy suggests that the UK’s target for a twentyfold increase in AI computing power could lead to a scenario where energy demand from digital infrastructure outstrips the growth of renewable energy capacity.
The UK’s decarbonisation efforts also hinge on electrifying transport and heating, sectors that have traditionally relied on fossil fuels. Transitioning from natural gas boilers to electric heat pumps and from combustion engine vehicles to electric vehicles will require substantial increases in electricity supply.
Addressing the challenges posed by AI’s energy demands will require more than just expanding renewable energy production. Enhancing the energy efficiency of AI systems and related technologies is crucial. Ensuring that the energy required for AI and other digital advancements is sustainably sourced, without compromising broader net zero goals, will necessitate a combination of government policy, technological innovation, and public awareness.
AI’s growing electricity needs could exacerbate competition for limited renewable energy resources, potentially increasing reliance on fossil fuels, especially during peak energy demand periods. If additional renewable capacity cannot be deployed quickly enough, the UK may face a situation where the electricity demand driven by AI increases overall emissions rather than reducing them.
The UK’s commitment to a twentyfold increase in public AI computing power by 2030 presents a formidable challenge for the country’s electricity system. Meeting this goal sustainably will require a careful balance between AI’s energy needs and the broader electrification goals necessary for achieving net zero. Without immediate and concerted efforts to expand renewable energy capacity and improve efficiency, the electricity demands of AI could hinder the transition to a sustainable, net zero future. The path forward will depend on strategic planning and collaboration across government, industry, and society to ensure that the benefits of AI do not come at the expense of environmental sustainability.
