Rich Kenny | Aug. 4, 2025

The promise of artificial intelligence (AI) is immense. It's poised to revolutionise industries, enhance public services, and drive economic growth, positioning the UK as a global leader in the field.

As we race toward this AI-powered future, a critical, and often hidden, environmental challenge is emerging: the significant and escalating water consumption of AI and data centers.

This issue is not a distant problem. As highlighted in a recent report authored by Interact MD Rich Kenny as Co-Chair of the GDSA Planetary Impact Working group and published on the Gov.uk website. The UK already faces a projected daily water deficit of nearly 5 billion litres by 2050. The rapid growth of AI and its associated water footprint, is set to add to this existing water stress, posing a potential threat to national water security.

The Three Pillars of AI’s Water Footprint

AI’s water consumption stems from three primary mechanisms, all of which are interconnected and contribute to its overall environmental impact:

Cooling Systems: Data centers generate vast amounts of heat. A common cooling method, evaporative cooling, uses immense volumes of fresh water, with about 80% of it being lost to evaporation. This process often uses potable (drinking quality) water and can significantly lower water levels in nearby areas, impacting aquatic ecosystems.

Electricity Generation: Powering these data centers and AI servers is a water-intensive process in itself. This "embedded" water footprint from electricity generation, particularly from fossil fuels, is frequently overlooked but can be as significant as the direct cooling needs. Global electricity consumption from data centers and AI servers is expected to more than double between 2022 and 2026, surpassing 1,000 TWh.

Chip Manufacturing: The production of the advanced semiconductor chips essential for AI hardware is another water-intensive process, requiring ultrapure water to rinse silicon wafers. While data center operators have less direct control over this "upstream" consumption, it’s a crucial component of AI’s overall water footprint that must be considered.

A Critical Policy Gap

Despite the scale of this challenge, a critical policy gap exists. Current national water resource plans, including those finalised in 2025, do not adequately account for the burgeoning demand from AI data centers. This systemic oversight risks intensifying water stress in already vulnerable regions of the UK, potentially leading to social and environmental conflicts and hindering economic development. The Environment Agency has even acknowledged that the future demand from new technologies like AI is "highly uncertain," making comprehensive planning difficult. This is compounded by a pervasive lack of transparency, with "no reliable data on the quantity of resources used by data centres," and only two-fifths of operators tracking their water usage.

Recommendations for a Sustainable AI Future

To proactively address these issues, the report outlines a series of strategic recommendations for the UK government, directly aligned with government objectives for sustainable AI development:

Mandate comprehensive, location-based reporting: Introduce legislation to require data centers above a specific power threshold (e.g., 1MW) to submit mandatory reports on their energy, water, and carbon emissions. This is crucial for transparency and accountability, as demonstrated by the EU's Common Union Rating Scheme.

Integrate AI water demand into national planning: Amend the upcoming National Framework for Water Resources (2025) and future Water Resources Management Plans to explicitly account for the projected water needs of AI data centers.

Implement water efficiency standards and incentives: Establish clear environmental sustainability requirements for new and existing data centers, focusing on reducing the use of potable water for cooling. Incentivise the adoption of advanced cooling technologies like liquid immersion and direct-to-chip cooling, which can reduce water usage by up to 52%.

Prioritise sustainable site selection: Mandate comprehensive water availability assessments for new data centers, particularly within government-backed "AI Growth Zones". The clustering of data centers in already water-stressed areas, such as the government’s first designated “AI Growth Zone” in Culham, Oxfordshire, risks exacerbating local water scarcity.

Lead by example: The government should strengthen its own Greening Government ICT Strategy to include explicit, quantifiable water efficiency targets for its own data centers and AI services.

By implementing these recommendations, the UK can ensure its ambitious pursuit of AI leadership is underpinned by robust environmental stewardship, safeguarding vital water resources for future generations and solidifying its position as a global leader in sustainable digital governance.

You can read the full report plus policy recomendations at Gov.UK

https://www.gov.uk/government/groups/government-digital-sustainability-alliance-gdsa