ARTIFICIAL Intelligence (AI) systems need to be powered. They require computer resources comprising hardware and infrastructure components.
These primarily include central processing units (CPUs), graphics processing units (GPUs), tensor processing units (TPUs), neural processing units (NPUs), as well as memory, storage, and networking.
Together, these components provide the processing power, data handling, and parallel-computation capabilities required to train, run and scale AI models and other intensive workloads.
These compute resources require energy-hungry data centres and AI factories.
A data centre is a physical facility that houses IT infrastructure, such as servers and storage, to manage, process and store data.
An AI Factory (which leverages data centres) is designed to develop, train, and deploy AI models at scale, focusing on computational resources and AI workflows.
Keep Reading
- Time running out for SA-based Zimbos
- Sally Mugabe renal unit disappears
- Epworth eyes town status
- Commodity price boom buoys GB
The power requirements of data centres and AI factories are at the gigawatt level. For example, a typical AI factory’s energy requirement is 2 GW (2000 MW), equivalent to that of a city such as San Francisco or a country such as Zimbabwe.
Furthermore, there is a demand for large volumes of fresh water for cooling (to prevent hardware overheating), putting pressure on the available water supply for human consumption, agriculture and other industrial purposes.
Hence, the large carbon footprint of AI systems primarily stems from the significant compute resources required for energy-intensive tasks such as training and deploying large language models (LLMs).
These tasks require high-performance hardware, such as GPUs and TPUs, which consume substantial electricity. For example, training a single LLM can emit hundreds of tonnes of carbon dioxide equivalent, comparable to the emissions of several cars over their lifetime.
Furthermore, as explained earlier, these models are trained and run in energy-hungry data centres, which must also be cooled.
Larger AI models require exponentially more energy. A model with billions of parameters demands far more computational power than a smaller one, resulting in a significantly larger carbon footprint.
Once trained, AI models continue to consume energy during deployment, particularly for real-time or large-scale applications.
Common examples include AI in content recommendation systems, search engines and real-time translation, which require continuous processing power.
The iterative nature of AI research, where models are trained and retrained to optimise performance, means that significant energy is expended in the development phase before a final model is deployed.
Hence, a country’s participation in driving the AI revolution (through data centres and AI factories) carries high environmental costs.
According to the International Energy Agency, total global electricity consumption by data centres could reach the level of Japan’s energy intake by 2026.
Another projection is that, in 2030, if all data centres worldwide were considered as one country, their overall energy demand would rank only fourth, behind China, the United States and India.
AI companies in highly-industrialised economies are even exploring the establishment of private nuclear power plants to meet their energy requirements.
Some countries are ramping up their fossil-fuel-driven power supplies to meet AI energy demands — a direct reversal of clean energy transition commitments.
United States (US) President Donald Trump revealed on January 23, 2025, while addressing the World Economic Forum that the US would have to double its annual electricity production to lead and drive the AI revolution.
That is the extent of the enormous energy demand exerted by the technology. President Trump intends to use the Executive Order (which he signed on January 20 2025) declaring a national energy emergency to address this challenge.
The legal instrument directs US agencies to utilise their statutory emergency powers to speed up the development and authorisation of energy projects. Unfortunately, with his slogan — “drill, baby, drill” — Trump’s AI energy plan will be anchored by boosting fossil fuel production to the detriment of global climate policies and regulations.
While many data centres are increasingly powered by renewable energy, non-renewable sources such as coal, natural gas, derived gas, crude oil and petroleum products are still often used, contributing to emissions.
Clearly, AI poses challenges to the global supply of adequate energy, threatens freshwater supplies and has the potential to worsen the climate change crisis.
Decarbonisation, energy generation
AI systems can be deployed to optimise energy generation, enhance grid management, increase renewable energy adoption, improve energy efficiency and expand energy access.
The technology can play a transformative role in making clean energy more accessible and sustainable. Thus, AI systems can drive innovation in energy management, grid optimisation, renewable energy forecasting, smart energy consumption, global decarbonisation and the expansion of access to clean energy in underserved regions.
This transformative role of AI supports the achievement of United Nations sustainable development goals (SDG) 7 (affordable and clean energy), which seeks to ensure universal access to affordable, reliable and modern energy services, increase the share of renewable energy in the global energy mix and improve energy efficiency.
Optimising energy generation
Efficient energy generation and distribution are essential for achieving universal access to clean, reliable energy. AI has the potential to enhance the performance of both traditional and renewable energy sources, while reducing waste and improving reliability.
Indeed, AI can enhance the efficacy of various clean energy sources such as solar, wind, geothermal, biofuels, and hydroelectric power.
In traditional power plants, AI-driven predictive maintenance systems monitor equipment conditions in real time, identifying potential issues before they lead to failures.
By analysing data from sensor-embedded turbines, generators and other machinery, AI algorithms can detect signs of wear and tear, allowing operators to conduct maintenance proactively.
Improving renewable energy
Renewable energy sources such as solar and wind are essential for reducing greenhouse gas emissions, but their intermittency presents challenges for a consistent energy supply.
AI-driven forecasting models have the potential to mitigate these challenges by accurately predicting energy generation from renewable sources, enabling operators to integrate them more effectively into the grid.
AI-based forecasting models use machine learning algorithms to analyse historical weather data, satellite imagery and sensor readings.
These models can predict energy generation levels hours, days, or even weeks in advance by identifying patterns in temperature, sunlight, wind speed and other factors.
This capability enables grid operators to plan for variability in renewable energy supply, ensuring a stable and reliable flow of electricity.
In solar energy, AI algorithms analyse satellite images and weather patterns to predict how much sunlight will reach solar panels. By providing accurate solar irradiance forecasts, AI helps optimise solar power generation and supports grid stability.
In wind energy, AI can predict changes in wind speed and direction, enabling wind farms to maximise energy production during favourable conditions.
This improves the efficiency of wind energy systems, reduces reliance on fossil fuels and makes renewable energy a more viable option for meeting energy demands.
Energy efficiency in industries
Energy efficiency is essential because reducing energy consumption minimises environmental impact and conserves resources. AI offers significant opportunities to improve energy efficiency in buildings, industrial facilities and transportation systems by enabling real-time monitoring, predictive analytics and automated energy management.
In buildings, AI-powered energy management systems monitor and adjust heating, ventilation and air conditioning (HVAC) systems based on occupancy patterns, weather conditions and user preferences.
These AI systems help reduce electricity bills, lower greenhouse gas emissions and promote sustainable building practices by optimising energy use.
In industrial facilities, AI enhances energy efficiency by monitoring machinery performance, optimising production processes and reducing resource use.
AI also optimises energy use in transport, particularly in electric vehicles (EVs) and public transport systems. AI algorithms analyse traffic patterns, energy consumption and charging needs to determine optimal routes and charging schedules for EVs, reducing energy use and improving battery efficiency.
In public transport, AI-powered systems optimise bus and train schedules to minimise idle time and fuel consumption, thereby promoting energy-efficient transport.
Access to clean energy
AI offers innovative solutions to expand clean energy access in remote and underserved areas by enabling decentralised energy systems, optimising micro grids and supporting off-grid renewable energy projects.
AI-driven microgrids (localised energy grids that operate independently from the central grid) provide a sustainable solution for remote communities with limited access to conventional energy infrastructure.
AI algorithms optimise microgrid performance by balancing energy generation, storage, and consumption based on real-time data.
By ensuring efficient energy distribution, AI-powered microgrids provide a reliable, affordable energy source for communities without access to traditional energy grids.
Smart energy consumption, demand
Smart energy consumption and demand response are essential for balancing supply and demand in energy systems, especially as more renewable energy sources are integrated into the grid.
AI supports these processes by enabling real-time monitoring, automated control and predictive analytics that help manage energy consumption more effectively.
In demand response programmes, AI algorithms analyse data on electricity prices, grid load and consumer behaviour to encourage users to shift energy consumption to off-peak hours.
AI also supports smart metering systems, which provide consumers with real-time information about their energy consumption patterns.
Smart meters use AI to analyse usage data, identify energy-saving opportunities and provide personalised recommendations to reduce energy use.
In addition, AI-enabled home energy management systems allow households to monitor and control appliances remotely, optimising energy use based on factors such as time of day, weather conditions, and electricity rates.
Way forward: Is AI a friend or foe?
It is instructive to acknowledge the complex interplay among AI, energy availability and decarbonisation. While the AI revolution demands enormous energy resources and has a large carbon footprint, the technology can also help create more efficient, sustainable and equitable energy systems globally.
Indeed, in the context of global energy supply and climate change mitigation, AI is a double-edged sword. The technology is both a problem and a solution!
The transformative technology is transforming industries by enhancing efficiencies and spurring innovation, but its expansion is imposing significant energy demands that could strain infrastructure and hinder climate progress.
In certain areas, increased data centre demand has driven up electricity costs for consumers and enterprises, underscoring the need for AI development to prioritise affordability and societal buy-in.
Globally, efforts are underway to address the complex nexus between AI, energy and decarbonisation. Strategies to reduce AI's carbon footprint are focusing on using renewable energy in data centres, developing more efficient algorithms and optimising models to reduce their size and energy requirements.
This is complemented by strategies for using AI to improve energy generation and demand management, while also driving decarbonisation.
The United Kingdom (UK) government's terms of reference for the independent Review of AI Deployment in Electricity Networks,published on December 16, 2025, outline a comprehensive assessment to address the growing complexities.
As electricity systems face challenges in forecasting, optimisation, flexibility and decarbonisation, AI technologies are seen as key to transforming planning, operations, and management, potentially enabling innovative approaches such as autonomous grids.
The UK review aims to map current and emerging AI applications with high-impact potential for the UK's energy system, identify technical, regulatory and data-related barriers and enablers, evaluate benefits for affordability, security, flexibility and decarbonisation alongside associated risks and deliver actionable recommendations for safe and rapid deployment.
This initiative supports broader objectives, such as clean power 2030 and the ambition to make Britain a clean energy superpower and is expected to deliver a final report by summer 2026.
The World Economic Forum (WEF) report titled From Paradox to Progress: A Net-Positive AI Energy Framework, published on
December 11, 2025, attempts to address this challenge presented by AI to the climate agenda and the efforts to resolve the global energy crisis.
The report correctly observes that without proactive measures, AI might inadvertently exacerbate system vulnerabilities and environmental risks, as evidenced by projections of soaring energy consumption.
By 2035, global data centre electricity use could exceed 1,200 terawatt-hours (TWh), up from 420 TWh in 2024."
It then proposes that achieving a net-positive AI energy outcome requires deliberate alignment among stakeholders to ensure that AI’s benefits in resource savings outweigh its overall use, fostering resilience and competitiveness.
The WEF framework for net-positive AI revolves around three action drivers — designing for efficiency, deploying for impact and shaping demand wisely — complemented by enablers like consumer education, ecosystem collaboration and transparent accountability.
This report draws from over 130 global use cases demonstrating tangible advantages, such as reduced costs, improved grid stability, and lower carbon dioxide (CO₂) emissions.
Challenges like transformer shortages and infrastructure constraints underscore the urgency of responsible scaling to avoid widening digital divides.
However, AI offers substantial opportunities, including optimising data centre cooling and enhancing heating, ventilation, and air conditioning (HVAC) systems, grids and industrial operations.
Net-positive AI energy means ensuring that the energy and resource savings enabled by AI outweigh its life cycle consumption – turning responsible scaling into a source of competitiveness and resilience."
Several companies and tech giants are proposing to establish AI data centres in space to address the escalating energy demands of terrestrial facilities, leveraging near-constant solar power in orbit for up to eight times greater efficiency and natural radiative cooling in a vacuum to eliminate water-intensive systems.
Google’s Project Suncatcher envisions constellations of solar- powered satellites equipped with TPUs and inter-satellite laser links, with prototype launches planned for 2027 to demonstrate scalable machine-learning compute.
Startup Starcloud, backed by Nvidia, launched an AI-equipped satellite with H100 GPUs into orbit in late 2025 and aims to build gigawatt-scale orbital facilities featuring massive 4-kilometre solar arrays.
SpaceX and Blue Origin are also advancing similar initiatives, potentially integrating compute into upgraded Starlink satellites or new orbital platforms.
In space, you get almost unlimited, low-cost renewable energy," highlighting the potential for dramatic reductions in costs and carbon emissions compared to Earth-based operations.
Yes, AI compute resources demand an inordinate amount of energy.
Embracing AI technology poses a threat to the climate change agenda and puts pressure on global energy supplies.
However, as discussed earlier, AI holds transformative potential for achieving affordable and clean energy (SDG 7) by enhancing energy generation, improving renewable energy forecasting, increasing energy efficiency, and expanding access to clean energy.
The technology provides innovative solutions to global energy challenges, from optimising smart grids and microgrids to promoting smart energy consumption and demand response.
Yes, in addressing the climate change agenda and achieving an adequate global energy supply, AI is both a friend and a foe.
However, creative and intentional deployment of AI, such as WEF’s Net-positive AI Energy Framework, and revolutionary innovations, such as building AI data centres in space, will make AI more of a friend than a foe.
This is an adapted excerpt from the book Deploying Artificial Intelligence to Achieve the UN Sustainable Development Goals:
Enablers, Drivers and Strategic Framework.
- Mutambara is the director and full professor of the Institute for the Future of Knowledge at the University of Johannesburg in South Africa.