Price of progress: AI is rapidly depleting the Earth’s water resources

Data centers powering AI are projected to consume 945 TWh of electricity annually by 2030, which is more than seven times Ukraine’s current energy usage.
AI’s environmental impact extends beyond just energy consumption to include significant water and land footprints due to cooling systems, infrastructure, and supply chains.
A United Nations University study suggests that AI could require as much water as the basic domestic needs of 1.3 billion people by the end of the decade.
The infrastructure for AI could expand to over 14,500 square kilometers, equivalent to twice the size of Jakarta, potentially worsening the environmental impact even with green energy projects.
Most of the energy used in AI is consumed during daily operations rather than during the initial training phase of AI models.
The environmental strain from AI is disproportionately felt in regions where new data centers are opening due to local power grid pressures and water shortages.
The AI industry is expected to generate about 2.5 million tons of electronic waste annually by 2030, burdening low-income countries that lack proper recycling facilities.
The current boom in AI is intensifying global inequality as over 90% of specialized AI computing power is dominated by the U.S. and China, while many other countries lack such infrastructure.