The advent of artificial intelligence (AI) technology has brought forward pressing discussions about its hidden environmental costs. While AI’s operations seem digitally seamless to users, the underlying processes, particularly water consumption in data centers, have drawn scrutiny. Extensive water usage, both directly in cooling systems and indirectly through power stations, is a significant feature of AI’s infrastructure footprint, leading to environmental considerations of its broader impact.
ETH
SOL
BSC
ROBINHOOD
USDT
AAPL
AI’s dependency on water is not a novel concern. Past discussions have consistently pointed to the environmental load carried by advancing technology, notably in data centers. Previously, organizations like the International Energy Agency flagged the substantial energy and water demand of these facilities, highlighting the challenge of balancing technological innovation with responsible resource use.
Why is the Water Consumption Estimate Important?
A study led by Alex de Vries-Gao sheds light on AI’s water footprint. The peer-reviewed research highlights an estimated consumption ranging from 312.5 to 764.6 billion liters during 2025, factoring in both direct use in data centers and water involved in electricity generation. Utilizing data from large corporations such as Google (NASDAQ:GOOGL), Meta (NASDAQ:META), and Apple (NASDAQ:AAPL), the study underscores significant variability and the need for transparency in reporting environmental usage.
What Does This Mean for Companies?
Understanding the distinction between direct and indirect water usage is essential for companies operating AI hardware. Large-scale AI operations tend to hide their extensive water needs due to the dominant share used in electricity production. The lack of specific reporting on AI operations deepens the ambiguity, making it challenging to precisely attribute water use to AI as opposed to other digital services.
Other challenges emerge when comparing AI’s water usage to consumer resources, such as bottled water. Although AI’s range of water consumption may seem comparable to the global bottled water industry’s scale, it’s crucial to recognize that methodologies differ. The environmental pressure induced by AI water usage depends heavily on geographic and climatic factors.
Can Innovation Alleviate Environmental Strain?
Companies like Microsoft (NASDAQ:MSFT) are exploring new designs that require less water for cooling AI systems, indicating possible directions for reducing environmental impacts without compromising technological progress. Still, the energy source behind these cooling systems remains an influencing factor on water footprint, hinting at a broader systemic issue beyond mere water usage.
Engagement in sustainable practices like optimizing AI operation locations according to environmental stressors, employing reclaimed water, and energy-efficient technological advancements can help mitigate AI’s environmental footprint. However, comprehensive solutions require a holistic view encompassing both policy and innovation.
Transparency in corporate reporting is vital to accurately assess the water consumption attributed to AI.
I recently wrote about why the 765 billion-litre figure deserves careful consideration
, emphasizing the significance of precision.
Companies could narrow the uncertainty by reporting where AI hardware operates
. The emphasis on comprehensive environmental accounting marks a critical step toward addressing these challenges.
