AI data centres could require as much water as the basic needs of 1.3 billion people by 2030 [1].
The scale of resource consumption raises urgent questions about the sustainability of the artificial intelligence boom. As these facilities expand, their impact on global water and energy supplies may create critical shortages in regions already facing climate instability.
Patrick Galey, head of investigations at Global Witness, highlighted these projections during a BBC Politics Live panel. Galey said the water demand of AI data centres could be equivalent to the basic needs of 1.3 billion people by 2030 [1]. This resource strain comes as the overall environmental footprint of data centres already rivals some of the world’s largest countries [2].
Industry growth is accelerating. Projections indicate that water, energy use, and pollution from AI data centres will double in four years as AI adoption grows [3]. The physical scale of new projects further illustrates this expansion — one newly approved AI data centre in the U.S. is the size of 15,840 Tesco Extras [4].
On the panel, political leaders debated how to balance these environmental costs against economic realities. Lucy Harris, a former Brexit Party MEP, said the priority of ordinary people is not the climate when it comes to the cost of living crisis.
Other panelists, including Labour MP Lola McEvoy, Green MP Siân Berry, and presenter Vicki Young, discussed whether the public should be more concerned about these infrastructure developments. The discussion focused on whether the current pace of AI deployment is outstripping the ability of governments to regulate its environmental toll.
“Water demand of AI data centres could be equivalent to the basic needs of 1.3 billion people by 2030.”
The tension between AI's rapid commercial scaling and its physical resource requirements suggests a looming conflict between technological ambition and environmental sustainability. If water and energy demands continue to double every few years, governments may be forced to choose between restricting AI growth or risking severe local ecological degradation and utility shortages.
