Artificial intelligence is no longer merely a supportive tool for work or research. Within just a few years, it has evolved into a massive industry with a very real environmental cost, one whose dimensions are only now coming into focus. At the forefront of these impacts is the growing pressure on water resources. This is a fact, not media exaggeration; it is the clear conclusion of scientific studies and credible international reports.

 

A research study conducted by scholars at Vrije Universiteit Amsterdam indicates that the global expansion of artificial intelligence technologies—particularly generative AI models and the data centres that support them—could lead to annual water consumption ranging between 312 and 764 billion litters by 2025. This is largely due to the direct use of water for cooling data centres, in addition to indirect water consumption associated with electricity generation.

 

An in-depth investigation by WIRED further confirms that most of this water use remains invisible to the average user. It is not limited to the water consumed inside data centres themselves but also includes vast quantities of water required to operate power plants that supply these facilities with electricity, what is known as the indirect water footprint.

 

The impact does not stop at water. The same study suggests that artificial intelligence could be responsible for between 32 and 80 million tons of carbon dioxide emissions annually, directly linking the water crisis to the broader challenges of climate change and energy consumption.

 

For its part, UNESCO, in its World Water Development Report, warns that unregulated digital transformation (including artificial intelligence) may create new pressures on water resources, particularly in countries already suffering from water scarcity. The report calls for integrating environmental and water considerations into national and global digital transformation policies.

 

The paradox is striking. Artificial intelligence is often promoted as a solution for improving water efficiency in agriculture and resource management, yet it is simultaneously used excessively for entertainment, leisure, and the mass production of consumer-driven content that adds little real value while steadily accumulating environmental pressure. Every unnecessary query and every intensive entertainment use mean more servers running, higher energy demand, and greater water consumption.

 

To avoid any misunderstanding, I am not opposed to artificial intelligence itself, but to its misuse. Technology is inherently neutral; it is the manner of its use that determines whether it becomes a tool for development or a new burden on already strained resources. Reports by the World Economic Forum emphasize the need to shift toward the concept of environmentally responsible AI, linking its expansion to sustainability standards rather than to profitability and speed alone.

 

In conclusion, despite its significant benefits, artificial intelligence must be used out of necessity, for productivity, knowledge, science, and improving quality of life, not as an unrestricted form of entertainment devoid of awareness or limits. Water security is no longer a marginal environmental concern; it is an existential issue. Any uncalculated depletion today will be paid for by humanity tomorrow.

 

In times of scarcity, rationalizing the use of technology becomes an ethical responsibility—just as essential as conserving water and energy.

 

Ahlam Nasser is a strategic planning and governance expert specializing in public sector reform, education policy, and sustainable development. She writes on public issues, including family and broader societal impacts.