AI's rapid expansion is creating a critical sustainability challenge that extends far beyond energy consumption: massive water demand from data center cooling operations is straining resources in regions already facing drought and water scarcity. While much of the focus on AI's environmental impact has centered on electricity use, a new analysis from investment firm BNP Paribas Asset Management reveals that water availability may become an equally pressing constraint as AI infrastructure scales globally.

Why Is Water Becoming AI's Most Overlooked Environmental Problem?

Data centers require enormous volumes of water for cooling systems, and as AI infrastructure expands into arid and drought-prone regions, competition for water between technology firms, agriculture, and local communities is intensifying. The challenge is particularly acute because many companies are building data centers in areas with limited water resources, driven by factors like lower costs and proximity to renewable energy sources.

For investors and companies deploying AI at scale, this creates multiple layers of risk. Physical risk emerges when water scarcity disrupts data center operations entirely. Drought risk materializes when data center water consumption contributes to regional shortages, causing operators to lose their "social license" to operate. Regulatory risk follows as municipalities impose water-use restrictions or pricing reforms that increase operational costs.

"Investors in datacentre operators, cloud service providers, and enabling infrastructure must now assess water exposure with the same seriousness historically reserved for mining or agriculture," noted Alex Bernhardt, Global Head of Sustainability Research at BNP Paribas Asset Management.

Alex Bernhardt, Global Head of Sustainability Research, BNP Paribas Asset Management

The water challenge represents a systemic risk that has largely escaped mainstream attention. While hyperscale tech operators increasingly procure renewable energy to address electricity concerns, the water dimension of AI infrastructure has received comparatively little scrutiny from regulators, investors, and the companies themselves.

How Can Investors and Companies Address AI's Water Footprint?

• Water Risk Assessment: Integrate water exposure analysis into due diligence processes for data center operators and cloud service providers, treating it with the same rigor applied to mining or agricultural companies.
• Regional Constraint Mapping: Evaluate where AI infrastructure is being deployed relative to local water availability, drought risk, and competing demand from agriculture and communities.
• Regulatory Engagement: Engage with policymakers and municipalities to develop water-use standards and pricing mechanisms that reflect true scarcity costs before crises force reactive restrictions.
• Technology Investment: Support development of water-efficient cooling technologies and alternative cooling methods that reduce data center water consumption.
• Transparency Requirements: Demand that technology firms disclose water usage metrics alongside energy consumption, enabling better risk assessment and accountability.

What Are the Broader Sustainability Implications of AI Expansion?

The water challenge is part of a larger sustainability equation surrounding AI. While water and energy constraints pose significant risks, AI also offers transformative opportunities to accelerate sustainability solutions. The technology can materially improve efficiency in energy, water, and material systems through smart grids, predictive maintenance, demand forecasting, and process optimization.

In industrial settings, AI-driven optimization can simultaneously cut energy use and input costs, strengthening both sustainability and competitiveness. Utilities and grid operators are adopting AI to manage distributed renewable energy systems more effectively. Logistics and transportation companies are using AI to optimize routing and fleet utilization, reducing emissions and fuel consumption.

Beyond operational efficiency, AI is accelerating discovery in critical sustainability domains. Machine learning is already being applied to materials science, battery chemistry, carbon capture technology, and drug development. By reducing the time and cost required to identify viable technologies, AI can compress innovation cycles that historically spanned decades.

For sustainable investors, the implication is clear: AI should be evaluated not only as a growth theme, but as a sustainability transition variable. The technology itself is neither inherently sustainable nor unsustainable; its net effect emerges from the interaction between markets, policy, and capital allocation decisions.

As electricity demand from AI is expected to double or triple by 2028 from 2024 levels in the United States, according to the US Department of Energy, the pressure on water resources will intensify alongside energy constraints. Companies and investors that proactively address water sustainability in AI infrastructure deployment will be better positioned to navigate regulatory changes, avoid operational disruptions, and maintain social license to operate in water-stressed regions.