Ecolab Study: Why AI’s Hidden Water Cost is a Risk to Growth

A report from Ecolab suggests the general public does not grasp the full extent of the water required to power AI.

The findings indicate a clear disparity between the awareness of AI’s energy demands and its substantial water consumption for cooling the data centres that are foundational to its operation.

The third annual Ecolab Watermark Study, which surveyed adults across fifteen countries, found that while most consumers understand AIs high power requirements, far fewer recognise the volume of water needed to regulate the temperature of its data centres.

According to the research, in the US, just 41% of consumers acknowledge water use in AI operations compared with 55% who recognise its high power consumption. This gap in public understanding is reportedly wider in the Asia Pacific and Latin America regions.

The scale of AI’s water consumption

These findings are particularly relevant as the global construction of data centres accelerates at a time of increasing water scarcity. The water demand is projected to create a large deficit in the coming years.

“By 2050, the world will have nearly 30% more people and require 47% more energy. Water demand will continue to grow rapidly – yet by 2030, the world already faces a projected 56% water deficit,“ says Christophe Beck, Chairman and CEO at Ecolab.

Christophe also highlights a fundamental difference between energy and water.

While energy can be generated from various sources, water is a finite resource, making it a necessary constraint on the expansion of the AI sector.

The study notes that just 20% of industrial wastewater is currently reused for data centre cooling systems, which could mean the majority of water consumed by AI is potable water.

“We see wastewater as an engineering flaw and a missed opportunity as a source for business growth,“ Christophe says.

Declining trust in corporate water management

The study also uncovered a growing trust deficit between the public and the corporate world on the issue of water stewardship.

While a majority of respondents agreed that businesses should invest in technologies to lessen AI’s impact on natural resources, there was doubt as to whether companies would manage water responsibly.

Fewer than half of consumers in the US (43%), Europe (44%), Asia Pacific (48%) and Latin America (48%) believe businesses are using smart water management strategies like reuse or recycling.

This sentiment could present a reputational and operational risk for AI companies, particularly those planning new data centre facilities.

Trust in government action on water reduction and recovery was marginally lower across all surveyed regions.

This disconnect is stark, given that 67% of consumers believe businesses and governments should prioritise efforts to lessen the impacts of climate change.

“Global consumers recognise smart water management is essential for a resilient future, and they expect businesses to lead with both transformative technologies and transparent action to make it a reality,“ explains Emilio Tenuta, CSO at Ecolab.

Emilio suggested that this presents an opportunity for businesses to deploy AI-driven water solutions that serve local communities while also fostering innovation.

AI as a solution to water scarcity

Despite the challenges outlined in the report, consumers remain optimistic about AI’s potential to create positive societal change.

The technology itself could provide solutions to the very problems it creates. Christophe suggests that through data analytics, real-time monitoring and temperature adjustments, AI could enable data centres “that use less water than a car wash".

This offers a pathway for the industry to mitigate its own environmental impact.

The research indicates a change in consumer purchasing preferences towards brands with responsible water practices. Increasing concern over water-related extreme weather events such as droughts, floods and wildfires is amplifying these anxieties.

This puts the onus on industries with large water footprints, like the AI sector, to demonstrate leadership. The study highlights that solving water challenges requires collaborative partnerships between businesses, governments and communities.

While agriculture remains the industry most identified with water consumption, the rapid growth of AI and data centres is an emerging concern that requires immediate attention.