Data centers, the global infrastructure that powers AI, could consume 945 terawatt-hours of electricity annually by 2030, nearly triple the combined annual electricity use of Pakistan, Bangladesh and Nigeria, countries that together are 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 data centers also carries a “water footprint” for cooling and power production, and a “land footprint” associated with power generation and supply chains.
Rethink how sustainability is measured
According to a new study from the United Nations University (UNU), AI-related water consumption could equal the annual basic household needs of 1.3 billion people by the end of the decade, while its territorial footprint could exceed 14,500 square kilometers, roughly double the size of the Jakarta metropolitan area.
In a data center, servers are high-performance computers that process and store data.
The report highlights a critical gap in how the environmental impact of AI is measured. Greenhouse gas emissions tend to be prioritized, particularly those related to training large models, but this approach overlooks other environmental costs.
Solutions considered “green” in one sense can worsen pressures in others, particularly in regions already facing resource shortages. For example, switching to certain renewable energy sources may reduce carbon emissions, but may significantly increase water consumption and land use.
Daily use of AI is the main culprit
Public debate has largely focused on the energy needed to train advanced AI models, but study finds that Daily use accounts for approximately 80 to 90 percent of total energy demand..
The scale is staggering: One widely used AI service is estimated to process around 2.5 billion messages per day, consuming hundreds of gigawatt-hours of electricity each year.
Energy usage also varies greatly depending on the task. Generating a single AI image can require more than a thousand times the energy of simple text classificationwhile video generation requires even greater resources.
Efficiency improvements alone are unlikely to offset these growing demands. The report points out the so-called rebound effect, in which lower costs and better performance drive higher usage and ultimately increase overall resource consumption.
Local charges, global benefits
The environmental impacts of AI infrastructure are not evenly distributed. While the benefits of technology are global, its costs are often concentrated in specific regions.
In some countries, data centers already account for a significant portion of national electricity consumption, putting pressure on energy systems. In others, Expanding facilities are drawing heavily on water supplies, sometimes amid drought conditions..
A server rack in a data center.
At the same time, the report warns of a The growing challenge of electronic wasteand AI infrastructure is projected to generate up to 2.5 million tonnes of e-waste annually by 2030. Much of this burden is likely to fall on low-income countries with limited capacity for its safe disposal.
The production of critical minerals needed for AI hardware also raises concerns about environmental degradation and social inequalities in extraction regions.
A growing digital and environmental divide
The expansion of AI infrastructure is also creating new disparities in access and influence. According to the report, more than 90 percent of computing capacity specialized in AI is concentrated in just two countries: the United States and China. At the same time, more than 150 countries lack significant national AI infrastructure.
This imbalance not only limits economic opportunities but also raises questions of environmental justice, as some countries bear the environmental costs without sharing the benefits of AI-driven growth.
Towards responsible AI
Despite the stark findings, the UNU researchers emphasize that the report is not an argument against AI itself. Rather, it demands urgent measures to ensure that technology is developed within planetary 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 urged to integrate AI infrastructure into energy, water and land use planning, while companies are encouraged to design systems that minimize resource consumption. Users also have a role to play by choosing lower-impact apps whenever possible.
Ultimately, the report argues that the future of AI will depend not only on technological innovation but also on governance decisions made today.