The AI boom has triggered a significant environmental crisis, with 2025's CO2 emissions equaling those of New York City, according to a recent report. This alarming revelation highlights the immense environmental impact of AI, which has been rapidly spreading globally. The research, published in the journal Patterns, estimates that AI-related water usage now surpasses the entire global demand for bottled water. The figures were compiled by Dutch academic Alex de Vries-Gao, founder of Digiconomist, who aimed to measure the specific impact of AI rather than datacentres in general. The study found that AI's carbon footprint in 2025 could reach 80 million tonnes, while its water usage could reach 765 billion litres, a third higher than previous estimates of all datacentre water use. This is a critical issue, as technology companies' environmental disclosures are often insufficient to assess their total impact, let alone isolate AI use. The IEA has reported that AI-focused datacentres draw as much electricity as power-thirsty aluminium smelters, and their electricity consumption is expected to more than double by 2030. This is a stark reminder that the public is footing the environmental bill for some of the richest companies on Earth, and the datacentre construction frenzy driven by generative AI is only getting started. The largest AI-focused datacentre in the UK, at a former coal power station site in Blyth, Northumberland, is expected to emit more than 180,000 tonnes of CO2 a year, equivalent to the amount produced by more than 24,000 homes. In India, where $30 billion is being invested in datacentres, there are growing concerns about the environmental impact of backup power, which could lead to a massive carbon liability. The question remains: is it fair that society is paying for these costs, while tech companies reap the benefits of this technology? The need for stricter transparency and accountability from technology companies is clear, as the environmental consequences of their actions are becoming increasingly apparent.
