By Tristan Fleck, Year 12
As artificial intelligence continues to expand, the environmental impact of AI data centres is becoming harder to ignore. While AI is often presented as a solution to sustainability challenges, the infrastructure that supports it requires large amounts of energy and clean water. One of the main concerns is the use of water for cooling AI data centres.
AI data centres generate extreme amounts of heat due to the constant processing of large volumes of data. To prevent overheating, many centres rely on water-based cooling systems. These systems use clean, drinkable water to absorb heat and keep servers running efficiently. In regions already facing water stress, this can place significant pressure on local water supplies.
The scale of water consumption is often underestimated. A single large data centre can use millions of liters of water per day, particularly during hot periods. This means that water which could be used for households, agriculture, or ecosystems is instead redirected to support digital services such as AI models, cloud storage, and online platforms.
Energy use is closely linked to this issue. When data centres rely on fossil fuel-based electricity, the environmental cost increases even further. High energy demand combined with heavy water use makes AI infrastructure more resource-intensive than many people realize. This raises questions about whether the rapid growth of AI is compatible with long-term sustainability goals.
However, solutions are beginning to emerge. Some companies are investing in alternative cooling methods such as air cooling, closed-loop systems, or the use of non-potable water. Others are placing data centres in cooler climates to reduce cooling demand or powering them with renewable energy sources. These approaches show that AI development does not have to come at the expense of essential resources.
AI has the potential to improve efficiency across many sectors, but its own environmental footprint must be addressed. Recognizing the water and energy demands of AI data centres is an important step toward responsible innovation. If technology is meant to support a sustainable future, should its hidden resource costs be more carefully managed?
