Environmental losses from AI threaten water, land and climate

Data centers, the global infrastructure that powers AI, could consume 945 terawatt-hours of electricity annually by 2030 – nearly three times the combined annual electricity use of Pakistan, Bangladesh and Nigeria, countries collectively home to more than 650 million people.

However, this is just the tip of the iceberg. In addition to the carbon footprint, each unit of electricity used by a data center also carries a “water footprint” for cooling and energy production, as well as a “land footprint” associated with power generation and supply chains.

Rethinking how sustainability is measured

According to the new one Study from the United Nations University (UNU), AI-related water consumption could equal the annual basic domestic needs of 1.3 billion people by the end of the decademeanwhile its land area can exceed 14,500 square kilometers – roughly twice the size of the Jakarta metropolitan area.

This report highlights critical gaps in how AI’s impact on the environment is measured. Greenhouse gas emissions, particularly those associated with large-scale model training, tend to be prioritized, but this approach ignores other environmental impacts.

Solutions that are considered “eco-friendly” can on the one hand exacerbate pressures in other countries, especially in regions already facing resource scarcity. For example, switching to certain renewable energy sources can reduce carbon emissions but can significantly increase water consumption and land use.

Everyday use of AI is the main cause

Public debate has largely centered on the energy required to train advanced AI models, but research has found this to be the case Daily use accounts for around 80 to 90 percent of total energy needs.

The scale is staggering: a widely used AI service is estimated to process around 2.5 billion requests per day, and consume hundreds of gigawatt-hours of electricity annually.

Energy use also varies greatly depending on the task. Generating a single AI image requires more than a thousand times more energy than simple text classificationwhile making videos requires greater resources.

Increased efficiency alone will not be able to keep up with this increase in demand. The report shows the so-called rebound effect, namely cost reductions and performance improvements that drive higher usage, ultimately increasing total resource consumption.

Local burden, global benefit

The environmental impacts of AI infrastructure are not spread evenly. While the benefits of this technology are global, the costs are often concentrated in specific regions.

In some countries, data centers already contribute a significant portion of national electricity consumption, putting pressure on the energy system. At another place, expanding facilities depleted water supplies, sometimes amid drought conditions.

At the same time, the report warns about a the growing challenge of e-wastewith AI infrastructure projected to generate up to 2.5 million tonnes of e-waste annually by 2030. Much of this burden will likely be borne by low-income countries with limited safe disposal capacity.

Production of critical minerals needed for AI hardware also raises concerns about environmental degradation and social inequality in extraction areas.

The digital and environmental divide is widening

The expansion of AI infrastructure is also creating new gaps in access and influence. According to the report, more than 90 percent of AI-specific computing capacity is concentrated in just two countries – United States and China. At the same time, more than 150 countries lack domestic AI infrastructure.

This imbalance not only limits economic opportunities but also raises questions about environmental justice, as some countries bear environmental impacts without sharing in the benefits of AI-driven growth.

Towards responsible AI

Despite the surprising findings, the UNU researchers stressed that the report is not an argument against AI itself. More correctly, it calls for immediate action to ensure that technology advances within the planet’s boundaries.

The study outlines a framework for a “responsible AI ecosystem,” built on principles including transparency, efficiency by design, equity, lifecycle responsibility, global cooperation, and sustainable use.

Governments are being urged to integrate AI infrastructure into energy, water and land use planning, while companies are being encouraged to design systems that minimize resource consumption. Users also have a role in choosing lower impact applications whenever possible.

Ultimately, the report argues that the future of AI depends not only on technological innovation but also on the governance choices made today.