Turkmenistan: Potential untapped—if it ever opens its doors
February 25, 2025
Why Georgia must push back against Emirati influence peddling
February 25, 2025
Artificial intelligence is transforming industries at an unprecedented pace, but its environmental footprint is expanding just as rapidly.
The growing demand for AI-driven technologies has placed increasing pressure on energy grids, water resources, and raw material supplies.
As AI systems grow more powerful and widespread, concerns over their sustainability are mounting, raising questions about whether the race for AI innovation can coexist with the urgent need for environmental responsibility.
AI’s soaring energy demands
The energy demands of AI are reaching staggering levels. The International Energy Agency reports that AI, data centres, and cryptocurrency operations consumed approximately 460 terawatt-hours of electricity in 2022, accounting for nearly two per cent of global electricity demand.
By 2027, AI alone could require between 85 and 134 terawatt-hours annually, an amount equivalent to the total electricity consumption of the Netherlands. The growth in AI infrastructure is already putting pressure on national energy supplies.
In the UK, the National Grid’s Group CEO, John Pettigrew, has warned that electricity demand from data centers could increase sixfold within the next decade, raising concerns that AI expansion may outstrip the country’s renewable energy capacity.
In Europe, demand for AI-related data centres continues to climb, with 601 megawatts of new capacity added in 2023, surpassing supply for the second time in five years. These figures suggest that AI’s expansion could significantly alter global energy consumption patterns, requiring new approaches to sustainability.
Water: The overlooked cost of AI growth
Beyond its power consumption, AI’s thirst for water is another growing concern. Data centers require vast amounts of water for cooling, often drawing from potable sources that could otherwise serve local populations.
In 2023, Google reported that 78 per cent of its global water withdrawals came from potable sources, heightening fears about water competition in regions where data centers operate. Some governments have already moved to limit the unchecked expansion of AI infrastructure.
Singapore, facing rising water and electricity constraints, imposed a moratorium on new data centers in 2019, which was subsequently adjusted. Meanwhile, the semiconductor industry, which supplies the hardware that powers AI, consumes an estimated 1.2 trillion liters of water annually.
The demand for ultrapure water in semiconductor fabrication is particularly concerning, with a single manufacturing facility requiring over 37 million litres daily.
In the UK, AI-driven water consumption is becoming an increasingly urgent issue. The UK Environment Agency warns that seven regions, including London, could face severe water stress by 2030, exacerbated by rising industrial and technological demands. As AI adoption accelerates, managing these water dependencies will be critical to ensuring sustainability.
The hidden environmental cost of AI hardware
AI’s environmental impact extends beyond electricity and water use, as the demand for specialised hardware introduces further sustainability challenges.
AI servers and processors require rare earth metals such as silicon, gallium, and tellurium, many of which are sourced from mining operations that disrupt ecosystems and deplete natural resources. The disposal of AI hardware presents yet another challenge, with rising levels of electronic waste linked to AI expansion.
The UK generated 23.9 kilograms of e-waste per capita in 2020, ranking second-highest globally. Projections suggest that AI-related e-waste could reach between 1.2 and five million tonnes by 2030 if disposal practices do not improve.
While recycling and reuse initiatives are expanding, many discarded AI components are still processed in developing regions where informal recycling operations lack adequate environmental and worker safety protections. The lifecycle of AI hardware is a growing sustainability issue that cannot be ignored.
Regulatory efforts and corporate sustainability initiatives
Governments and regulatory bodies are beginning to take steps toward mitigating AI’s environmental impact. The United Nations Environment Programme advocates AI-specific climate impact reporting, urging companies to disclose energy, water, and material consumption to enhance transparency.
Meanwhile, discussions within the G20 and the Organisation for Economic Co-operation and Development are exploring the introduction of carbon pricing for AI workloads to encourage energy-efficient models. The European Union’s Energy Efficiency Directive is also pushing for stricter efficiency and carbon reduction requirements for data centres across the continent.
Tech companies are making their pledges to address AI’s sustainability challenges. Google, Microsoft, and Amazon have all committed to making their AI operations carbon-neutral by 2030 or 2040. Hardware manufacturers such as NVIDIA and IBM are developing energy-efficient AI chips designed to reduce power consumption by up to 25-fold over previous generations.
Some firms are exploring alternative cooling methods, with Microsoft piloting direct-to-chip and immersion cooling technologies that could significantly reduce AI’s reliance on traditional water-cooled systems.
Despite these commitments, concerns remain about whether the measures being taken are sufficient. Some investigations suggest that major tech companies may be underreporting the emissions of their data centers, raising questions about the transparency of corporate sustainability pledges. More comprehensive data and independent oversight will be necessary to ensure that AI’s environmental impact is accurately assessed and managed.
The future of AI: Can innovation align with sustainability?Â
As AI adoption continues to accelerate, balancing its resource demands with sustainability will be one of the defining challenges of the coming decade. Without intervention, AI’s increasing electricity and water consumption could undermine energy security, intensify water shortages, and accelerate climate change.
A coordinated global effort will be required to ensure AI’s future remains sustainable. Stronger regulatory frameworks, increased investment in green computing innovations, and greater corporate accountability will all be necessary to align AI’s expansion with environmental stewardship.
The future of AI is at a crossroads. The question is no longer whether AI will continue to grow, but whether its development can proceed without further straining the planet’s resources. The decisions made today will determine whether AI becomes a tool for sustainable progress or an unchecked driver of environmental degradation. The challenge now is to ensure that AI’s potential is realized in a way that respects and preserves the world’s finite resources.
Read the complete article at The Hidden Cost of AI: Energy, Water, and the Sustainability Challenge
This content has been produced in collaboration with a partner organisation through our Global Visibility Programme. Our programme helps companies and organisations boost their digital presence and strengthen the thought leadership of their experts. Find out more here.
