AI’s Expanding Energy and Water Demand Raises Global Sustainability Concerns

A United Nations‑affiliated report warns that artificial‑intelligence systems are set to become major consumers of electricity, water and land. In 2025 AI‑related data centres used about 448 terawatt‑hours (TWh) of electricity, and projections indicate demand could rise to 945 TWh by 2030 – roughly the current total electricity consumption of the African continent and enough to rank the sector as the world’s sixth‑largest electricity consumer.

The same analysis estimates a water footprint of 9.3 trillion litres per year for AI data‑centre operations by 2030, equivalent to the basic annual water needs of 1.3 billion people in sub‑Saharan Africa, and land use of about 14 500 km², comparable to twice the size of Greater Jakarta. Current estimates also show that a single ChatGPT query can consume up to 60 times the energy of a standard web search, while generating a video can use the same power as 200 000 spam‑filter classifications.

Report authors stress that most existing assessments focus only on carbon emissions, overlooking water and land impacts, and call for a transparent, efficient and responsible AI ecosystem to mitigate these unintended environmental burdens.