The United Nations has released a report projecting that by 2030, artificial intelligence could consume 3% of the world’s electricity and use more water for cooling than humanity’s annual drinking water needs, raising significant concerns about sustainability and equity.
- AI could consume 3% of global electricity by 2030
- Water usage for AI cooling may exceed global human drinking water needs
- Most AI infrastructure concentrated in US and China, widening digital divide
What happened
The United Nations published a report warning of the environmental costs associated with widespread AI adoption. By 2030, AI's electricity consumption could double from current levels to account for 3% of global power usage, while the water required for cooling these systems might surpass the annual drinking water needs of the entire human population.
The report also points out that improvements in AI efficiency may trigger the Jevons paradox, where technology that saves resources increases overall demand due to lower operational costs. As AI models become more affordable and widespread, total consumption of electricity and water could escalate, potentially negating the benefits of energy-saving advancements.
Why it matters
This surge in AI resource consumption raises sustainability concerns as data centers already consume electricity equivalent to that of Saudi Arabia, the world’s 11th largest electricity user. A doubling in consumption would require planting 6.7 billion trees over ten years to offset carbon emissions and vast new demands on water and land resources.
Beyond environmental strain, the report uncovers structural inequities in AI infrastructure deployment. Currently, just 32 countries have AI-specific cloud infrastructure, with 90% located in the US and China. This concentration risks deepening the digital divide and transferring environmental burdens, such as e-waste and mineral extraction, to less developed countries.
What to watch next
The report advocates for responsible AI development that integrates environmental disclosures, transparency, and lifecycle governance from mineral sourcing to disposal. It urges global cooperation to balance AI capabilities with environmental stewardship, emphasizing sustainable development and justice.
In countries like Australia and New Zealand, AI strategies focus on AI adoption in public sectors but currently lack mandatory environmental impact reporting. Moving forward, regulators and policymakers should consider embedding sustainability requirements and comprehensive energy monitoring into AI frameworks to mitigate the sector’s growing environmental footprint.