Article is AI-generated for the purpose of filling in entries. Its not paraphrased or polishing my own writing. This project, per se, is just a demonstration.
A new scientific analysis released in late 2025 suggests that the global environmental impact of artificial intelligence (AI) systems is far larger than many people realize. According to the study, the total carbon emissions associated with powering and cooling AI systems in 2025 could be on par with the yearly emissions of a major city — specifically comparable to New York City. At the same time, AI’s water usage may rival the entire world’s annual consumption of bottled water. :contentReference[oaicite:1]{index=1}
Carbon Footprint: As Big as a City
The researchers estimated that the carbon dioxide emissions from AI computing — mostly from data centers and the electricity needed to power them — could reach tens of millions of tonnes in 2025. Emissions at that scale have been compared with the annual carbon footprint of New York City, a metropolis of over 8 million people. :contentReference[oaicite:2]{index=2}
Water Footprint: Thirsty Machines
AI systems require a lot of water, especially for keeping servers cool in data centers. In hot, high‑performance facilities, water is used to regulate temperature and prevent overheating. The study suggests that the total annual water usage of AI in 2025 could be in the same ballpark as the global bottled water market’s annual consumption — hundreds of billions of liters of water. :contentReference[oaicite:3]{index=3}
Why This Matters
This research highlights a growing challenge: while AI technologies — including large language models and generative systems — are becoming increasingly useful in everyday life, they also rely on massive computing infrastructure that consumes energy and water at scale. As demand for AI services rises, so too do the environmental costs. :contentReference[oaicite:4]{index=4}
Caveats: It’s Still Hard to Measure Precisely
Experts note that measuring the exact environmental footprint of AI is difficult. Tech companies don’t always separate AI‑specific usage from other computing in their environmental reports, and data centers often power many different tasks at once. As a result, researchers must estimate AI’s impact indirectly, which means the numbers are approximate — but they still provide a useful sense of scale. :contentReference[oaicite:5]{index=5}
What This Means Going Forward
The findings have sparked discussion about the sustainability of AI development. Some scientists and policy experts are calling for greater transparency from tech companies about energy and water use, as well as more investment in greener data center technologies, renewable energy sources, and cooling systems that consume less water. :contentReference[oaicite:6]{index=6}
Ultimately, understanding the full ecological cost of AI could help guide smarter decisions about how these powerful technologies are designed, powered, and regulated — balancing innovation with environmental responsibility.