2025 Study Reveals AI’s Massive Water and Carbon Footprint

A newly published peer-reviewed study has shed light on the significant environmental impact of artificial intelligence systems in 2025, revealing levels of resource consumption that rival entire global industries.

According to the research, AI technologies consumed an estimated 312.5 to 764.6 billion litres of water-equivalent over the year — a figure comparable to the global bottled water industry’s annual usage. Much of this demand comes from the need to cool vast data centres that power advanced models such as ChatGPT. These facilities rely heavily on evaporative cooling systems to manage the intense heat generated by high-volume computational workloads.

The study also highlights AI’s growing carbon footprint. Researchers estimate emissions ranging between 32.6 and 79.7 million tonnes of CO₂, roughly on par with New York City’s annual emissions from energy use and transport combined.

As AI adoption accelerates across industries, experts are warning that its environmental cost cannot be ignored. They are calling for increased investment in renewable energy, more efficient data centre design, and advanced cooling technologies to reduce the strain.

The findings underline a growing contradiction: while AI is often promoted as a tool to tackle climate challenges, its unchecked expansion risks making it a major contributor to the problem.

#MaltaDaily