Water crisis in Britain: AI data centers drain the country's resources
The UK faces a serious challenge: the rapid scaling of artificial intelligence infrastructure threatens the country's water security. Industry analysts are sounding the alarm: existing water supply capacities are not designed to meet the demands of modern data centers.
Why is water the new oil for AI?
Data centers consume colossal amounts of water to cool servers. Cooling towers and liquid cooling systems operate around the clock, and growing electricity needs only exacerbate the situation. According to expert estimates, one large data center can consume up to 3 million liters of water per day — comparable to the peak consumption of a small town with 3,500 households.
Forecast errors and hidden risks
An analysis of government documents shows that the UK's AI development strategies completely ignore the water aspect. Not a single program — neither the AI growth zones nor the "AI for Science" plan — even mentions water. This is a fatal mistake, as current water consumption forecasts clearly exclude data centers from calculations.
The most alarming situation has developed in the Slough region, where Affinity Water serves about 125 data centers, with a hose pipe ban already in effect across most of the area. Local operators are requesting up to 3 million liters of water per day per facility, creating critical strain on infrastructure.
Alarming figures
Current regulations require data centers to use exclusively drinking water, leading to a daily consumption of about 6.6 million liters. If the government fulfills its plan to triple computing capacity by 2030, this figure will soar to 19.8 million liters per day. Meanwhile, building new reservoirs could take up to 15 years, and tightening leak and consumption standards has nearly exhausted its potential.
"Any further growth in the number of data centers in water-scarce regions may prove impossible without a radical policy overhaul. The only way out is either to increase water abstraction from the environment, which would harm it, or to build new infrastructure, but that will take years," industry analysts emphasize.
Significantly, similar problems have already forced New York authorities to impose a moratorium on the construction of large data centers until a corresponding law is passed. In Texas, data center operators are even bypassing public environmental oversight. Britain risks repeating these mistakes if it does not take preventive measures.
My conclusion: The AI industry finds itself in a paradoxical situation — a technology designed to optimize resources is itself becoming their main consumer. Without the implementation of closed-loop water supply systems and a transition to alternative cooling methods, the water crisis could become a serious brake on the development of artificial intelligence in the UK. Investors should closely monitor regulatory changes in this sector — they could radically alter the economics of data centers.