It has been claimed that due to the AI boom, as much carbon dioxide will be released into the atmosphere in 2025 as is emitted in the entire New York City.
Research estimates global environmental impact of rapidly expanding technology published on Wednesday, which also found that AI-related water use now exceeds the entire demand for global bottled water.
The figures have been compiled by Dutch academic Alex de Vries-Gao, founder of DigiconomistA company that researches the unintended consequences of digital trends. They claimed they were the first attempt to measure the specific impact of artificial intelligence rather than the datacenter in general as the use of chatbots like OpenAI’s ChatGPT and Google’s Gemini increases in 2025.
Statistics show that greenhouse gas emissions estimated using AI are now also equivalent to more than 8% of global aviation emissions. Their study used technology companies’ own reporting and called for stricter requirements to be more transparent about their climate impact.
“The environmental cost of this is absolutely huge,” he said. “Right now society is paying these costs, not the technology companies. The question is: Is that fair? If they’re reaping the benefits of this technology, why shouldn’t they pay some of the costs?”
De Vries-Gao found that the 2025 carbon footprint of AI systems could be up to 80m tonnes, while water used could reach 765bn litres. He said this was the first time the water impact of AI had been estimated and revealed that AI water use alone was one third higher than previous estimates of all datacenter water use.
These data have been published in the academic journal Patterns. International Energy Agency (IEA) Said Launching this year, AI-focused datacenters draw the same amount of power as power-thirsty aluminum smelters, and datacenter power consumption is expected to more than double by 2030.
“This is further evidence that the public is footing the environmental bill for some of the richest companies on earth,” said Donald Campbell, director of advocacy at Foxglove, a UK non-profit that campaigns for fairness in tech. “Worse, this is likely just the tip of the iceberg. The datacenter construction frenzy, driven by generative AI, is just getting started.
“Just one of these new ‘hyperscale’ facilities could generate climate emissions equivalent to many international airports. And in the UK alone, there are an estimated 100-200 in the planning system.”
The IEA reports that each of the largest AI-centric datacenters being built today will consume as much electricity as 2 million homes, with the US (45%) accounting for the largest share of datacenter electricity consumption, followed by China (25%) and Europe (15%).
The largest datacentre in the UK is being planned on a former coal power station site in Blyth, Northumberland, which is expected to emit more than 180,000 tonnes of CO2 per year during full operation – equivalent to the amount produced by more than 24,000 homes.
In India, where $30bn (£22.5bn) is being invested in datacentres, there are growing concerns that the National Grid’s lack of reliability will mean building huge diesel generator farms for backup power, consultancy KPMG reported this week. called “A huge…carbon liability”.
De Vries-Gao said technology companies’ environmental disclosures are often inadequate to assess the total datacenter impact, let alone isolate the use of AI. He said that when Google recently reported on the impact of its Gemini AI, it did not account for the water used in generating the electricity needed to power it.
Google reported that in 2024 it managed to reduce energy emissions from its datacenters by 12% due to new clean energy sources, but it Said Achieving its climate targets this summer was “now more complex and challenging at every level from local to global” and “a key challenge is the slower deployment than required of carbon-free energy technologies at scale”.
Google has been contacted for comment.
