The artificial intelligence revolution consuming global attention runs on a foundation of profound environmental destruction and labor exploitation, according to UN researchers whose findings challenge tech industry sustainability claims.
The production of critical minerals essential to AI data centers—lithium, cobalt, copper, and rare earth elements—extracted primarily from Chile, the Democratic Republic of the Congo, Peru, Bolivia, Argentina, and Zambia, consumed an estimated 456 billion liters of water in 2024 alone—equivalent to the annual domestic water needs of approximately 62 million people in sub-Saharan Africa.
In Chile's Salar de Atacama, mining accounts for up to 65% of regional water use, competing directly with agriculture and fragile desert ecosystems in one of Earth's driest regions. The water depletion occurs as tech companies tout their environmental commitments while expanding AI infrastructure at unprecedented rates.
Rare earth mineral production generates approximately 2,000 metric tons of waste per metric ton of usable material, creating toxic wastewater containing heavy metals, acids, and radioactive residues. Rivers near cobalt and copper mines have become so acidic that communities cannot safely drink from them, and fish stocks have collapsed entirely.
The human cost proves equally devastating. Communities near mining sites report elevated rates of skin diseases, gastrointestinal illness, and reproductive health problems. In the Democratic Republic of the Congo, maternity wards near mining operations report significantly higher rates of birth defects. Chile's Antofagasta region has the highest cancer mortality in the country, with lung cancer rates nearly three times the national average.
Thousands of children work in artisanal cobalt mines in the DRC, often without protective equipment against cobalt dust and other hazardous materials. The exploitation occurs at the supply chain foundation for technologies marketed as solutions to climate change and human advancement.
In climate policy, as across environmental challenges, urgency must meet solutions—science demands action, but despair achieves nothing. The UN findings reveal that AI's environmental footprint extends far beyond data center energy consumption to encompass resource extraction systems causing immediate, measurable harm to vulnerable populations.
"Critical minerals are essential to advancing sustainability," the researchers conclude. "But if cleaner technologies are built in ways that result in polluted rivers, sick children and dispossessed communities, the transition will fall short of its promise."
The research challenges fundamental assumptions about technological progress. While Silicon Valley executives discuss AI's potential to solve climate change, the infrastructure enabling these systems depletes water resources communities need for survival, poisons ecosystems, and relies on labor conditions approaching modern slavery.
Environmental justice advocates emphasize that genuine sustainability requires addressing extraction conditions, not merely optimizing data center efficiency. The minerals fueling AI development cannot be separated from the communities and ecosystems sacrificed to obtain them.
Tech companies have increasingly announced commitments to renewable energy for data centers and carbon neutrality targets. Yet these pledges rarely address supply chain impacts or resource extraction conditions. The disconnect between sustainability rhetoric and extraction reality grows more pronounced as AI development accelerates.
The UN findings suggest that without fundamental changes to critical mineral sourcing—including enforceable labor standards, environmental protections, and equitable benefit-sharing with affected communities—the AI boom perpetuates colonial extraction patterns under technological advancement's guise.
Several mining regions lack basic regulatory enforcement, enabling practices that would be illegal in countries where AI systems are developed and deployed. The geographic separation between harm and benefit reflects broader patterns in climate and environmental injustice.
Researchers emphasize that technological solutions exist—improved recycling of rare earth elements, alternative battery chemistries reducing cobalt dependence, and extraction methods minimizing water use. Implementation requires political will and corporate accountability currently absent from industry practices.
The question facing the AI industry is whether revolutionary technology will be built on revolutionary ethics, or whether it will replicate the exploitation patterns characterizing earlier industrial transformations. The answer will determine not only AI's environmental legacy but its fundamental legitimacy as a force for human advancement.
