Labor & Employment Law Daily Wrap Up, AI NEWS—United Nations University report outlines AI environmental costs, posits management framework, (Jun 8, 2026)

By Joe Cox, J.D.

The report focuses not only on the massive projected costs of AI but on some important considerations for governance in managing the technology.

United National University’s (UNU) Institute for Water, Environment and Health recently released a 56-page report titled Environmental Cost of AI’s Energy Use: Carbon, Water and Land Footprints that details both the growing cost of artificial intelligence’s resource footprint and potential governmental concerns in attempting to handle the anticipated future growth of the technology.

The report notes that given the resources implicit in the development of AI, its continued growth and societal cost “involves physical infrastructure and supply chains, including data centers, chips, electricity generation, cooling systems, water withdrawals, land occupation, critical minerals, and eventual e-waste.” In assessing and forecasting future solutions, the report and an associated press release noted that UNU looked to move beyond the “low-carbon” discussion that has dominated AI planning, as “low-carbon” planning might not be “low-water” or “low-land” and could merely result in hiding trade-offs and shifting environmental burdens.

Projected AI growth and resource cost. The report projects the global AI market to expand from roughly $189 billion USD to in 2023 to nearly $5 trillion by 2033, with global AI expenditure projected to reach $2.5 billion in 2026. The divide between the haves and have-nots is emphasized by AI, with 60 percent of jobs in advanced economies incorporating AI, versus 26 percent in lower-income countries.

Also notable was the amount of energy consumed and to be consumed by GPT-4, for instance, its training energy was noted to be equivalent to the annual residential electricity consumption of over 460,000 people in Sub-Saharan Africa, with GPT-4’s carbon footprint requiring sequestration capacity of the amount of trees in 105 Hyde Parks in London.

The UNU report further discussed the energy costs of AI in use. The report noted that a single short AI video can draw as much electricity as 200,000 computer spam classifications, and that the energy required to generate a typical AI image is enough to power a 10-watt LED bulb for 17 minutes.

Data center electricity consumption. Additionally, data centers were assessed as a resource cost. The report notes that if 2025 data centers’ electricity use was considered as a country, it would have ranked 11th globally in energy consumption. That rate is projected to push up to sixth globally by 2030 based on current patterns. Otherwise stated, the energy consumed by data centers in 2025 was enough to supply the annual residential electricity needs of the entire population of Sub-Saharan Africa, 1.3 billion people, for 2.6 years.

Future planning and governance. Aside from projections of future growth, the report focused on managing the future of AI. UNU advocates for a responsible AI ecosystem centering around six key principles: transparency, efficiency by design, equity and environmental justice, lifecycle responsibility, global cooperation, and sustainable use. A key component across most of these key principles was open and accurate reporting on energy use, largely so that governments could accurately integrate AI infrastructure into energy system planning, carbon accounting, water governance and land use permitting.

The report further emphasized that siting decisions should further be open and transparent so that innovation can advance without unduly burdening vulnerable communities, particularly in environmentally stressed regions. Data which is open and comparable would help governments and communities make more accurate and forward-looking plans in handling the anticipated spread of the AI footprint.

News: IndustryNewsTrends AINews Reports