Asger Risom
A landmark UN report warns that the rapid global expansion of artificial intelligence is creating an unsustainable environmental footprint, threatening massive water consumption and millions of tonnes of e-waste by 2030.
Artificial intelligence is driving a massive expansion in global resource consumption that threatens to severely strain the planet’s natural ecosystems. A new landmark study warns that the technology’s environmental footprint is scaling at a pace that could rapidly outstrip global sustainability targets.
Staggering demands on power and water
The physical infrastructure required to operate foundational AI models is creating unprecedented demands on local utilities.
According to a comprehensive study released by UN News, the global data centers powering AI could consume 945 terawatt-hours of electricity annually by 2030.
Alongside this immense power draw, the report highlights that the water required to cool these facilities and generate electricity could equal the basic domestic water needs of 1.3 billion people by the end of the decade.
The hidden toll of daily usage
While public debate has largely centered on the carbon emissions generated by training massive models, everyday commercial usage is proving to be the primary environmental driver.
A detailed breakdown by the United Nations University reveals that day-to-day model inferencing accounts for roughly 80% to 90% of total AI energy demand.
For example, researchers estimate that generating a single high-resolution AI video clip can require hundreds of times more electricity than processing simple text classifications, fundamentally shifting how energy budgets must be calculated.
E-waste and widening global divides
The rapid expansion of AI hardware is also creating severe long-term disposal challenges that disproportionately impact developing economies.
The United Nations researchers warn that global AI infrastructure is projected to generate up to 2.5 million tonnes of toxic electronic waste annually by 2030.
Much of this hazardous burden is expected to fall heavily on lower-income countries that lack the industrial capacity for safe recycling, deepening the environmental divide between the nations developing AI and those bearing its physical costs.
