Phil Robotham | World Water Day | AI’s unquenchable thirst

While the mammoth energy requirement for AI and Large Language models has been widely publicised, World Water Day – celebrated on 22 March 2025 – provides a good opportunity to shine a light on a lesser-covered environmental impact of AI – its water footprint.

A study by the University of Massachusetts Amherst found that training a single generative AI model can consume as much as 284,000 litres of water. That's equivalent to the amount of water an average person would consume over the course of 27 years.

While AI is a digital technology, it depends on physical processes with real environmental impact. Datacentres consume huge amounts of energy and water and making semiconductors requires incredibly sophisticated equipment and facilities.

According to a Forbes article, AI's projected water usage could hit 6.6 billion m³ by 2027, signalling a need to tackle its water footprint.

The water footprint of datacentres is driven by both the water consumed for the electricity generation and the water consumed for cooling. This is a significant problem because the use of AI is becoming increasingly widespread across a range of industries – especially since recent innovations by DeepSeek which are fast tracking the democratisation of AI. The demand for datacentres and the associated water consumption is only going to increase.

But there are ways for data centres to reduce their operational water footprint

The electrical energy going into a datacentre must eventually be ejected as waste heat through a cooling system. Datacentre cooling accounts for 33-40% of overall datacentre energy usage and consumes hundreds of billions of litres of fresh water per year.

On the positive side, there are steps that can, and are being taken to decrease the water requirements for data centres. Recently there have been significant advances in datacentre cooling technologies that have improved efficiency and reduced costs.

The location of the datacentres itself can make a huge difference too. Given that around 70% of freshwater is used in food production, building data centres in areas with preexisting water scarcity will have indirect impacts for agriculture, and may exacerbate the higher incidence of droughts that we are already seeing. By building datacentres in locations with abundant renewable power and/or cooler climates, significant emissions savings can be made.

In addition to this, there are various innovative methods for repurposing the heat generated by datacentres. Among these is using the excess warmth to bolster district heating systems, whereby high temperature water can be channelled into households and buildings.

Alternatively, this thermal energy may be harnessed for agricultural applications, such as providing year-round heating for greenhouses, or for heating water in fish farms and public swimming pools.

Sustainable food and water systems require massive investment

Our food and water system will be exposed to unprecedented challenges over the next 30 years. Aside from the environmental impact of the AI revolution, the need to reduce greenhouse gas emissions and adapt to a changing climate whilst improving biodiversity, reducing pollution and waste, and encouraging healthier diets globally has never been greater.

Early estimates suggest that this structural shift will require around $30trillion of capital reallocation by 2050. This is a significant step change in investment rates which we believe will create opportunities for investors across the entire food and water value chain.

*Phil Robotham, Head of Intermediary in South Africa at global investment manager, Schroders.