With the rapid rise of artificial intelligence, the world’s data centres have become some of the largest consumers of water. Every day, these massive facilities use millions of litres for cooling, raising concerns that the growing demand could further strain already scarce water resources.

AI and its hidden water cost

Few users realise that each AI chatbot response, on average, consumes about 500 millilitres of clean drinking water. While half a litre may seem negligible, the number becomes alarming when multiplied by the billions of queries asked daily worldwide.

Research shows that a single data centre uses at least 200 million litres of water every day. That equals about nine billion litres per month, or over 100 billion litres annually — enough to provide drinking water to millions of people in a large city for months.

How data centres use water

Thousands of powerful servers inside data centres run continuously, generating massive amounts of heat. To prevent overheating and maintain performance, water-based cooling systems are widely used.

The world’s largest AI data hubs are based in countries like the United States, China, Ireland, Spain, Singapore, and the Netherlands. Ironically, many of these nations are already grappling with water scarcity crises, raising questions about sustainability.

Impact on agriculture, drinking water

Experts caution that this rising trend is putting direct pressure on agriculture and drinking water reserves. As global temperatures rise and populations grow, diverting such vast amounts of clean water to data centres could worsen shortages in vulnerable regions.

Environmentalists have called for urgent international regulations to prevent unchecked water use by the tech industry. Without global action, they warn, communities could face deeper crises in access to safe drinking water.

Calls for global action

Campaigners argue that international laws must be introduced to stop water wastage linked to AI-driven technologies. If not regulated, they say, the race for technological advancement could come at the cost of human survival.