AI effect: Global water consumption could surge 11x by 2028
What's the story
Artificial intelligence (AI) data centers are likely to lead to a massive surge in global water consumption, a Morgan Stanley report has revealed. The study predicts that by 2028, these facilities could consume some 1,068 billion liters of water annually for cooling systems and electricity generation. This would be an astounding 11-fold increase from estimates made in 2024.
Water usage
Understanding the 3 'scopes' of AI's water footprint
The report highlights that while the cooling of data centers is a well-known source of water consumption, the indirect usage through electricity generation is often overlooked. The full water footprint of AI covers three "scopes." Scope one refers to on-site cooling or direct use of water by data centers to prevent servers from overheating, while scope two relates to electricity generation and its indirect use in powering these facilities.
Manufacturing impact
Semiconductor manufacturing also contributes to water usage
The third scope pertains to semiconductor manufacturing, the water-heavy process of producing specialized chips for AI. Semiconductor facilities can use up to five million gallons of ultrapure water every day. The report's estimates are based on various factors such as assumptions about water intensity and consumption factors, cooling technologies, and regional energy mix.
Future projections
Adoption of more efficient solutions
Depending on the adoption of more efficient solutions, annual water consumption could range between 637 billion liters and a staggering 1,485 billion liters. This highlights the potential variability in future water usage patterns as operations become more efficient.
Regional risks
Data centers are increasingly located in areas facing risks
The report also notes that while AI's overall global water consumption may seem small, its impact is highly localized. Over half of the world's top data center hubs are located in areas that are already facing a risk due to threats from drought, flooding, and declining water quality. This trend is seen in many secondary markets and emerging data center hubs as well.