UN report warns AI could soon use 3% of world’s electricity and more water than we need to drink

A recent report from the United Nations has raised significant alarms regarding the environmental implications of artificial intelligence (AI), projecting that by 2030, AI could consume as much as 3% of the world’s electricity and use more water for cooling than what is required for global drinking needs. This report challenges the common assumption that advancements in AI technology will lead to reduced resource consumption, warning instead of a potential increase in demand due to the phenomenon known as the "Jevons paradox."

The Jevons paradox, named after economist William Stanley Jevons, posits that as technological improvements make a resource more efficient, the overall consumption of that resource can actually rise. This counterintuitive outcome is attributed to lower operational costs that encourage expanded use. In the context of AI, as models become cheaper and more efficient, the report suggests that this will likely lead to greater overall usage, thus negating any potential benefits from increased efficiency.

Projected Environmental Impact

According to the UN report, the energy consumption of AI is set to double by 2030, equating to the annual electricity use of a country the size of Saudi Arabia, which is the 11th largest electricity consumer in the world. This surge in energy demand would generate a carbon footprint that would require approximately 6.7 billion trees to be planted over a decade to offset. Furthermore, the report estimates that data centers, which are crucial for AI operations, will require around 9.3 trillion liters of water for cooling purposes, a volume that is staggering compared to the annual drinking water needs of the global population.

The report also highlights a concerning trend regarding the geographical distribution of AI infrastructure. Currently, only 32 nations host AI-specific cloud infrastructure, with a staggering 90% of this capacity located in the United States and China. This concentration raises concerns about a widening digital divide, where countries that develop and control AI technologies reap the benefits, while those that merely consume them face disproportionate environmental burdens linked to mineral extraction and electronic waste.

The Need for Responsible AI

To mitigate these issues, the UN report advocates for a roadmap centered on responsible AI use, emphasizing principles such as transparency, efficiency by design, equity, and global cooperation. The report underscores the importance of full value-chain governance, which encompasses everything from the sourcing of minerals used in AI hardware to the recycling and safe disposal of electronic waste. It calls for a dual focus on enhancing AI capabilities while simultaneously protecting the natural environment.

This approach necessitates that environmental disclosures become a standard part of AI development, ensuring that both the models and the tasks they perform are evaluated for their environmental impact. Furthermore, it suggests that projections of AI demand should be integrated into broader climate and energy planning efforts.

As countries around the world, including Aotearoa New Zealand and Australia, begin to adopt national AI strategies, the report warns that current regulatory frameworks may not adequately address the environmental costs associated with AI. For instance, while New Zealand’s national AI strategy is informed by values-based principles, it lacks requirements for environmental disclosures. Similarly, Australia’s national AI plan, which aims to improve public services, may overlook the growing environmental implications of AI technology.

The report emphasizes that the natural environment is foundational to economic stability, cultural integrity, and overall well-being. Therefore, it argues for a fundamental shift in how we approach AI innovation, advocating for a sustainable technology future that prioritizes environmental stewardship alongside technological advancement. As AI continues to evolve, it is crucial that creators and technologists heed these warnings and integrate responsible practices into their work to ensure a balanced relationship between technology and the environment.