The rapid global expansion of artificial intelligence is set to drive a 129% increase in water demand across the artificial intelligence (AI) value chain by 2050, according to new research by Xylem Inc. (Washington, D.C.; www.xylem.com) and Global Water Intelligence (GWI; Oxford, U.K.; www.globalwaterintel.com). Watering the New Economy: Managing the Impacts of the AI Revolution provides the most comprehensive assessment to date of how AI is reshaping global water use. It also highlights practical ways to meet that demand while strengthening water security for communities.
Spending on AI is forecast to approach $2 trillion in 2026 [1] accelerating infrastructure build-out worldwide. By the middle of the century, this expansion is projected to add 30 trillion liters of water demand, annually, fueled by a surge in power generation (~54% of the increase), semiconductor fabrication (~42%), and data center expansion (~4%).
Crucially, the report finds that this additional demand does not have to result in competition between industry and communities. Today, global water systems process roughly 320 trillion liters of wastewater each year and lose an estimated 100 trillion liters through aging infrastructure – volumes that, if recovered through targeted investment, could substantially offset AI’s projected growth in water use and reduce strain on freshwater resources.
The report calls for a coordinated “water transition” centered on water reuse, digital infrastructure, and cross-sector partnerships – to ensure water is a resilient foundation for economic growth, rather than a constraint on innovation.
“AI is placing new demands on water supplies, but many of the tools needed to address the challenge already exist,” said Matthew Pine, Xylem’s president and CEO. “Advanced treatment technologies, for example, allow us to recycle water rather than waste it. Digital systems can help better manage supply in real time, reducing water lost to leaks. It’s time for a water transition built on targeted investment and collaboration between industry, utilities, and governments to ensure water systems can support both growth and community resilience.”
The analysis also highlights rising geographic risk. Nearly 40% of existing data centers are already located in regions experiencing high water stress [2], and future growth is expected to concentrate in similarly constrained watersheds. This underscores the need for localized, basin-level solutions, even as global efficiency gains create headroom for growth.
“Our projections examine water use across the full AI value chain – from chip fabrication and data center operations to indirect demand from power generation – and assess how technology choices shape future demand,” said Christopher Gasson, CEO, Global Water Intelligence. “The greatest pressure points emerge in semiconductor manufacturing and in fast-growing data center hubs in the United States, East Asia, and South Asia. In these regions, expanded wastewater reuse, leakage reduction, and targeted infrastructure investment can fully offset future growth.”
The report points to early examples of how collaboration can deliver shared benefits. In Mexico, for example, utilities in Mexico City and Monterrey worked together with Amazon and Xylem on smart pipeline upgrades to reduce leaks. A system of sensors and advanced analytics will save more than 1.3 billion liters of water a year and help improve the reliability of water supplies for residents.
[1] Beneath the surface: Water stress in data centers | S&P Global
Key findings of Watering the New Economy: Managing the impacts of the AI revolution include the following:
- In 2025, water use associated with the AI value chain – including data centers, semiconductor manufacturing and supporting power generation – is estimated at approximately 23.7 km³ (23.7 trillion liters) per year.
- This represents around 3.7% of total global industrial water withdrawals, underscoring that AI’s water footprint is material but smaller than many legacy industries.
- AI-related water demand is projected to increase by approximately 129% by 2050, rising to more than 54 km³ (over 54 trillion liters) annually:
- Data centers: +272% water demand growth by 2050
- Semiconductor manufacturing: +613% water demand growth by 2050
- Power generation: +18% water demand growth by 2050
- Nearly 40% of existing data centers are in regions facing high or extremely high water stress.
- Thirty minutes of intensive AI use requires just over 600 ml of water, significantly less than many previous estimates.
- The largest share comes from power generation, followed by water embedded in chip manufacturing and onsite data center cooling.
- The projected increase in AI-related water demand could be substantially offset through improved water efficiency and reuse, reducing pressure on freshwater resources:
- Up to 160 km³ per year is potentially available from wastewater reuse.
- Up to 100 km³ per year could be recovered through reducing losses in aging water distribution networks.
- While these measures could meet aggregate global demand, local water stress challenges will persist and require targeted solutions.