A new UN report warns that the environmental footprint of artificial intelligence is expanding faster than the rules meant to manage it. Researchers say water consumption, land use, and carbon emissions from AI data centers are rising sharply, while governance frameworks lag behind. The report calls for broader metrics beyond just carbon accounting to capture the full resource burden of AI infrastructure. For carbon markets and climate policy, this matters because AI's energy demand is already straining grids and competing with renewable energy projects. Data center buildout is driving new power purchase agreements, but also raising questions about water scarcity in drought-prone regions. Policymakers are being urged to set standards that cover the full lifecycle impact of AI, not just operational emissions.
