The environmental costs of artificial intelligence threaten water, land and climate

Data centers powering AI systems could consume up to 945 terawatt-hours of electricity annually by 2030, which is three times the total annual electricity consumption of Pakistan, Bangladesh, and Nigeria combined, countries with a population of more than 650 million.
However, this is just the tip of the iceberg. In addition to the carbon footprint, each unit of electricity consumed by data centers has a “water footprint” for cooling and energy production, and a “land footprint” associated with power generation and supply chains.
Rethinking the sustainability scale
According to a new study by the United Nations University, water consumption associated with artificial intelligence may equal the basic annual household needs of 1.3 billion people by the end of the current decade, while its land footprint may exceed 14,500 square kilometers, equivalent to about five times the area of Cairo Governorate.
The report highlights a fundamental gap in how the environmental impact of artificial intelligence is measured, as the focus is often on greenhouse gas emissions, especially those associated with training large models, while this approach overlooks other environmental costs.
Solutions that are considered “green” from one angle may exacerbate pressures in other aspects, especially in areas that already suffer from resource scarcity. For example, switching to certain renewable energy sources may reduce carbon emissions, but may cause a significant increase in water consumption and land use.
In a data center, servers are high-performance computers that process and store data.
Everyday use of artificial intelligence
Public debate has largely focused on the energy needed to train advanced AI models, but the study showed that daily use accounts for between 80 and 90 percent of total energy demand, and the numbers are staggering.
It is estimated that one widespread AI service processes about 2.5 billion routing commands per day, consuming hundreds of gigawatt-hours of electricity annually.
Power consumption also varies greatly depending on the task. Generating a single image using artificial intelligence may require more than a thousand times more energy than it takes to classify simple text, while generating video clips requires much greater resources.
Improvements in efficiency alone are unlikely to offset this increased demand. The report referred to what is known as the rebound effect, where lower costs and improved performance lead to increased utilization rates, which ultimately leads to an overall increase in resource consumption.
Local burdens and global benefits
The environmental impacts of AI infrastructure are not evenly distributed. While the benefits of this technology are global, its costs are often concentrated in specific regions.
In some countries, data centers already account for a large share of national electricity consumption, putting pressure on energy systems. In other countries, expanding facilities are significantly depleting water resources, sometimes even under drought conditions.
At the same time, the report warns of a growing challenge in e-waste, with AI infrastructure expected to generate up to 2.5 million tons of e-waste annually by 2030. Much of this burden is likely to fall on low-income countries that lack sufficient capacity to safely dispose of this waste.
The processes of producing vital minerals needed for artificial intelligence devices also raise concerns about environmental degradation and social injustice in the areas where extraction operations take place.
Internet servers in a data center.
The digital and environmental divide is widening
The expansion of AI infrastructure is also creating new inequalities in access and influence. According to the report, more than 90 percent of the computing capabilities specialized in artificial intelligence are concentrated in only two countries: the United States and China. In contrast, more than 150 countries lack tangible domestic AI infrastructure.
This imbalance not only limits economic opportunities, but also raises questions about environmental justice, as some countries bear the environmental costs without enjoying a share of the benefits of AI-driven growth.
Towards responsible artificial intelligence
Despite these stark findings, UNU researchers stressed that the report does not aim to oppose artificial intelligence per se, but rather calls for urgent action to ensure that this technology develops within the planet’s ecological limits.
The study proposed a framework for a “responsible artificial intelligence system,” based on principles that include transparency, efficiency from the design stage, fairness, responsibility throughout the life cycle, global cooperation, and sustainable use.
It urged governments to integrate AI infrastructure into energy, water and land use plans, and encouraged companies to design systems that reduce resource consumption. Users also have a role to play, which is to choose applications with a lower environmental impact whenever possible.
In conclusion, the report concluded that the future of artificial intelligence will depend not only on technological innovation, but also on the governance choices made today.
- For more: Follow Khaleejion 24 Arabic, Khaleejion 24 English, Khaleejion 24 Live, and for social media follow us on Facebook and Twitter



