Perhaps the word most associated with “AI” right now is “bubble,” but it is increasingly understood that it is much more than just an economic time bomb. It also carries significant public health risks. Beyond the release of pollutants, the demand for large amounts of clean water from AI data centers could reduce sanitation in nearby communities, worsen gastrointestinal illnesses, and put additional strain on local health infrastructure.
AI consumes significant amounts of energy and becomes increasingly dependent on water
Generative AI is artificial intelligence that can generate new text, photos, code, etc., and is already pervasive in the lives of most people around the world. ChatGPT alone is reported to receive approximately 1 billion queries per day, indicating huge demand on an individual level.
However, this is just the tip of the iceberg. Companies like Google, Apple, and Microsoft are now incorporating AI into their core products. Applications that leverage search results are rapidly moving toward adopting AI as the new standard for algorithms. Whether it’s shopping on Amazon or booking flights or hotels, searches now use AI and require more energy. As an example, a single AI-powered Google search is estimated to use up to 30 times more energy than the standard version.
Graphics processing units (GPUs) are the industry’s current answer to this. These are chips that require energy and generate heat. Thousands of small cores on a GPU enable parallel processing of large, repetitive calculations performed by AI algorithms, but a single chip can use up to 700 watts. This means that just three chips can use almost the same amount of energy as a domestic electric oven.
The large amounts of heat generated by data centers is cooled by up to hundreds of thousands of gallons of fresh water each day. With thousands of heat generating chips stacked next to each other, a simple fan won’t help. Instead, water is pumped or immersed between and around the chips to avoid overheating the system. A recent report from the UK Government Digital Sustainability Alliance predicts that AI will increase global water use from 1.1 billion cubic meters to 6.6 billion cubic meters by 2027.
Some companies are trying to use seawater for cooling. However, fresh water continues to be widely used for cooling in many facilities. Water recycling is an option, but it’s not easy. Some companies use “closed loop systems” to reduce the total amount of water required. Nevertheless, dust and minerals collected during cooling can deteriorate water quality over time, requiring treatment or replacement.
How AI-driven water scarcity threatens public health
Locating data centers in areas where water is already scarce can quickly lead to health care burdens before contamination becomes a problem. In 2023, Microsoft reported that 41% of its water withdrawals came from water-stressed regions. Meanwhile, Google said 15% of its water consumption occurs in areas with severe water scarcity. Amazon does not provide comparable numbers.
It is well established that water scarcity is correlated with infectious diseases, malnutrition and poor hygiene. Most such studies focus on already impoverished areas, but these are often the very places where data centers are planned to be built. Moreover, the root cause remains the same. The lack of fresh water for local residents has forced households to prioritize drinking and cooking over hand-washing, eating, and bathing. Naturally, this also reduces the water available for cleaning your living space.
The World Health Organization recognizes that unsafe drinking water and poor sanitation contribute to the spread of diseases such as cholera, other diarrheal diseases, and various other pathogens. To make matters worse, some diseases associated with water scarcity can be transmitted from person to person, increasing the risk of sustained community transmission.
The health burden on children is particularly worrying because infection and death are more common than in adults. In fact, 84 percent of the global burden of diarrheal diseases is borne by children under the age of five, and infections with diarrheal pathogens are associated with cognitive consequences from early childhood onwards.
Although it is too early to establish a direct causal relationship between AI data centers and water-related diseases, the known facts make this a serious concern. It has been established that AI data centers can significantly deplete local water supplies. We also know that areas with poor access to water are at increased risk of gastrointestinal illness and other illnesses.
Claiming that AI data centers directly cause gastrointestinal illness is unlikely to hold much weight. But the warning signs are becoming increasingly difficult to ignore. If the risk is foreseeable and severe, governments should not wait until people start dying before implementing preventive policies.
Marginalized communities are already reporting contaminated water
In Newton County, Georgia, where Meta built an AI data center, residents are reporting discolored water filling their faucets, which they blame on the facility. Similarly, in Fayette County, residents have reported sediment in their bodies of water, which is believed to have coincided with the construction of a nearby data center. Another report from the state of California suggests that a data center planned along the San Francisco Bay in Bayview-Hunters Point raises concerns that it will further increase the environmental burden on an already polluted community. In all of these cases, the local population includes large numbers of blacks and African Americans, a pattern that raises environmental justice concerns.
Accumulated residues can cause effects ranging from acute gastrointestinal illness to chronic diseases such as cancer. Microbial contamination can cause poisoning and acute illness, while chemical residues involve long-term harm and often act as a slow, invisible threat.
Data center plans in African countries such as Nigeria, Egypt and South Africa raise further questions about who will bear the brunt of the environmental impacts and whether affected communities will receive adequate protection and support. The situation is uncertain due to weak regulatory oversight in some of these countries. In many cases, serious community impacts may go unreported at all.
Why corporate water commitments and regulations matter
Only 0.5% of the planet has fresh water, but water is needed for more than just data centers. It is also required for power plants that generate electricity. Water usage is not just a data center issue, but an AI supply chain issue, as manufacturing of chips and interconnects similarly requires water.
Many companies are committed to sustainability, and some even claim to be “pure water producers” or “water positive.” Even if such a goal is achieved, which remains questionable, it must benefit the communities where the water is extracted. Providing more water to wealthy areas while supplies dry up in places like Newton County may meet corporate accounting standards, but local residents will still suffer the consequences.
To fulfill their ethical obligations to their citizens, governments must rapidly keep up with the pace of AI expansion and data center construction. A healthy population is more productive, and reducing the burden on public health can reduce government spending while supporting development. More fundamentally, there is a collective moral obligation to build a sustainable future for generations to come by protecting water security and avoiding environmental catastrophe. This starts with laws that require transparent reporting of corporate water use and enforce meaningful standards for sustainable management. Regulation must prioritize human well-being over short-term growth in extractive technologies. As with climate change, unchecked innovation risks further harm to both people and the planet.
The views expressed in this article are the author’s own and do not necessarily reflect the editorial stance of Al Jazeera.
