UNU study warns AI’s energy, water and land costs

UN University (UNU) published a study on AI’s environmental footprint.

• Main announcement: The UNU study finds that AI infrastructure and data centers could consume 945 terawatt-hours of electricity annually by 2030, with AI-related water consumption equal to the basic annual domestic needs of 1.3 billion people by the end of the decade, a land footprint that may exceed 14,500 square kilometres, and e-waste projected up to 2.5 million tonnes annually by 2030. The report also states day-to-day usage accounts for roughly 80 to 90 per cent of total energy demand and cites a widely used AI service processing ~2.5 billion prompts per day (consuming hundreds of gigawatt-hours per year).

• Background and recommendations: The study highlights that more than 90 per cent of AI-specialised computing capacity is concentrated in the United States and China, while over 150 nations lack significant domestic AI infrastructure; it calls for a “responsible AI ecosystem” built on transparency, efficiency by design, equity, lifecycle responsibility, global cooperation and sustainable use, and urges governments to integrate AI infrastructure into energy, water and land-use planning and companies to design systems that minimise resource consumption.