The recent research has shed light on AI-induced environmental costs characterized by high CO2 emissions and water consumption.
The research was conducted by Dutch academic Alex de Vries-Gao, the founder of Digiconomist, a company that maps the consequences of technology-based trends.
The study is considered a first-of-its-kind effort to analyze the broad-spectrum effects of artificial intelligence, including chatbots, Gemini and ChatGPT, on Earth.
This is evident from collected figures published in the magazine PatternsAI-powered greenhouse gas emissions have reached levels equivalent to over 8 percent of global aviation emissions.
Moreover, the carbon footprint caused by the large-scale deployment of AI systems could reach 80 million tons by 2025. Water consumption could also reach 765 billion liters.
Alex de Vries-Gao said: “The environmental costs of this are quite large in absolute terms. Right now society is paying for these costs, not the tech companies.”
“The question is: is that fair? If they are reaping the benefits of this technology, why shouldn’t they have to pay some of the costs?” he added.
Earlier this year, the International Energy Agency also explained the costs of AI-powered data centers in the form of high electricity consumption. By 2030, consumption will more than double.
“Worse, this is likely just the tip of the iceberg. The data center building frenzy powered by generative AI is just beginning. Just one of these new ‘hyperscale’ facilities can generate climate emissions equivalent to those of several international airports,” said Donald Campbell, director of advocacy at Foxglove.
According to IEA reports, these data centers have the capacity to consume as much power as 2 million homes.
Of all countries, the US is responsible for the largest consumption share of 45 percent. China and Europe hold 25 and 15 percent of the shares respectively.
