The report details how AI's value chain, from semiconductor manufacturing to data center cooling, consumes substantial amounts of water. It notes that indirect water usage, particularly in power generation and chip production, often surpasses the direct consumption at data centers themselves. Policymakers have reportedly overlooked the resource implications of AI's swift growth, with data center water usage projected to double by 2030.

Community concerns regarding water access and electricity demands are also identified as a constraint on AI development, leading to delays or modifications in large data center projects. The report points to opportunities for AI to enhance water management, citing UAE projects like MaiWater that integrate AI with renewable energy and water planning.

The report offers five recommendations for future AI development: fostering international cooperation on responsible AI water use, integrating AI infrastructure into water and energy planning, mandating transparent resource-use reporting, promoting low-water and low-energy technologies, and embedding water and climate risk into corporate strategies. The upcoming UN Water Conference in Abu Dhabi is highlighted as a key opportunity to establish a shared global agenda for responsible AI development.