en.Wedoany.com Reported - A new study released by the United Nations University (UN University) points out that the global race to scale up artificial intelligence is placing a broad environmental burden on energy, water, land, mineral, and waste systems, extending beyond the carbon footprint. The report emphasizes that the current focus on greenhouse gas emissions, particularly those associated with training large models, is too narrow a measurement approach, overlooking other pressure points that have equally substantial impacts.
The report notes that the environmental cost of AI should not merely be seen as a byproduct of digital productivity tools; it is becoming a physical infrastructure sector, increasingly involving energy planning, water rights, land use, and waste management. Research shows that energy demand driven by daily AI usage accounts for 80% to 90% of total demand. A widely used AI service processes approximately 2.5 billion prompts per day, consuming hundreds of gigawatt-hours of electricity annually. Energy consumption varies greatly between tasks; generating a single AI image can require over a thousand times more energy than text classification, while video generation consumes even more. The report also highlights a rebound effect—cheaper and faster systems lead to higher usage, so even if individual tasks become more efficient, total demand may still rise.
The report estimates that by 2030, data centers could consume 945 terawatt-hours of electricity annually, nearly three times the total annual electricity consumption of Pakistan, Bangladesh, and Nigeria combined (with a total population exceeding 650 million). Meanwhile, by the end of this decade, AI-related water consumption could be equivalent to the basic annual household water needs of 1.3 billion people. Its potential land footprint could exceed 14,500 square kilometers. These pressures are unevenly distributed across regions: in water-stressed areas, AI infrastructure may compete with households, agriculture, and industry for water; in power-constrained markets, data centers could add burden to grids already under pressure from electrification and climate adaptation. The challenge of electronic waste is also growing; by 2030, AI infrastructure could generate up to 2.5 million tons of e-waste annually, with much of the burden potentially falling on low- and middle-income countries with limited safe disposal capacity. Additionally, the critical mineral supply chain on which AI hardware depends may cause environmental damage and social injustice in mining regions.
The report links the growth of AI infrastructure to global inequality. Over 90% of AI-specific computing power is concentrated in the United States and China, while more than 150 countries lack significant domestic AI infrastructure. This imbalance limits economic participation and raises environmental justice concerns—some regions may bear the extraction, energy, water, or waste impacts of AI without equal access to its economic benefits. The report calls for strengthened governance to align AI development with planetary boundaries and proposes a responsible AI ecosystem based on transparency, design efficiency, fairness, lifecycle responsibility, global cooperation, and sustainable use. It urges governments to integrate AI infrastructure into energy, water, and land-use planning, encourages companies to reduce resource consumption from the design stage, and advises users to choose lower-impact applications where possible.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
