how safe are these AI trends for us, and the environment that sustains us While that’s fair. Read on to find out.
The ordinary people who upload their pictures to recreate a 1980s retro look for a moment of joy might not be aware that ChatGPT and Gemini can save their pictures. AI itself claims that “on personal ChatGPT accounts, content may be used to improve/train models unless you opt out. Unless turned off, data control options allow OpenAI and Google to save user data for training purposes. Most large-scale AI systems run in data centres, including those managed by cloud service providers. However, these facilities can have a significant environmental impact. In fact, in Thane and Visakhapatnam, residents have staged some of India’s first organised protests against large data centres , raising concerns that these projects could further strain local land, power and water resources. Data centres require larger amounts of water to cool them down and generate electricity.
As reported by Al Jazeera, a study by Cambridge researchers found that land surface temperatures around AI data centres increase by an average of 2°C (3.6°F), with some areas experiencing rises of up to 9°C (16.2°F). Data centres consumed around 415 terawatt hours (TWh) of electricity in 2024, accounting for roughly 1.5% of global electricity consumption, as per Al Jazeera, according to the International Energy Agency (IEA). This demand has grown by about 15% annually over the past five years and is expected to nearly double to 945 TWh by 2030. A study published in Water Research estimated that AI’s global water footprint could rise to 4.2–6.6 billion cubic metres a year by 2027, while noting that around two-thirds of data centres built after 2022 are located in water-stressed regions.
This is called the “data heat island effect”. A case-study by University of Oxford labelled data centres as “water guzzling” and “energy eaters”.
