The big tech industry’s claims about the climate benefits of artificial intelligence (AI) are largely unproven and unsubstantiated, according to a new report from a coalition of climate advocacy and accountability groups.
The report found that only 26 percent of the climate claims made by big tech companies cited published academic research, with 36 percent citing no evidence whatsoever.
The analysis is the first of its kind to assess climate claims from major AI developers like Google and Microsoft, as well as from independent institutions like the International Energy Agency (IEA). The prevailing narrative in the tech industry has been that the benefits of AI will more than offset the massive increase in emissions expected from new data centres.
Many of the claims made by big tech conflate the climate benefits of ‘traditional’ AI – machine learning tools designed to streamline specific tasks, like email spam filters, which have relatively low carbon emissions – and ‘generative’ AI chatbots.
The latter – including platforms such as ChatGPT – are a huge driver of increased emissions, mainly through the construction and operation of new fossil fuel-powered data centres, which deliver the massive computing capacity needed to service generative AI. A query in ChatGPT requires about 10 times the computing power of a standard Google search.
Traditional AI can help to combat climate change by processing vast, complex datasets to identify patterns and optimise systems.
However, the report concluded that: “At no point did this search or analysis uncover examples in which generative systems were leading to a material, verifiable and substantial level of emissions reductions.”
“By muddling these two types of AI into one umbrella term, purported climate solutions are coupled to extreme pollution and presented as a package deal,” the report stated.
Ketan Joshi, an independent climate and energy analyst, who authored the report, added: “It appears tech companies are using vagueness about what happens within energy-hogging data centres to greenwash a planet-wrecking expansion”.
The average data centre uses enough energy to power roughly 5,000 UK homes, and between 11 million and 19 million litres of water per day, the same as a town of 30,000 to 50,000 people. There are 480 data centres in the UK, with another 100 planned for the next five years.
In order to deliver the energy required for these data centres, the big tech industry has been extending the life of coal power plants that were otherwise scheduled for closure, as DeSmog has reported. Across the world, big tech companies are also attempting to power their data centres using gas, and in so doing throwing a lifeline to fossil fuel infrastructure. Canadian gas pipeline builder TC Energy announced in 2024 that the AI boom was pushing gas demand to a “record high”.
“An environmentally destructive heavy industry has emerged in the blink of an eye,” the report said. “The rapid expansion of data centres to drive the deployment of technologies marketed as ‘artificial intelligence’ rescues prospects for fossil fuels by boosting demand and triggering panicked deployment of new fossil infrastructure. This corrosive trend is often justified on the grounds that ‘AI’ will ultimately undo these sins by serving as a net benefit to climate action.”
Joshi added: “The promises of planet-saving tech remain hollow, while AI data centres breathe life into coal and gas every day. These claims of climate benefit are unjustified and overhyped, and could cover up irreversible damage being done to communities and society.”
The International Energy Agency forecasts that the energy requirements of data centres will quadruple by 2030 – to nearly as much energy expenditure as Japan. The UK’s National Energy System Operator (NESO) says that, by 2030, data centres will guzzle 7 percent of Britain’s entire energy output.
Researchers from Loughborough University have also warned that if the same rate of AI growth continues through to 2033, data centres across the globe will demand more power than all the electricity currently generated on Earth.
In addition, AI is being used in ways that actively harm the environment – including via ‘traditional’ machine learning systems. The report added that: “AI is used by fossil fuel companies to streamline exploration and increase extraction. Similarly, applications such as the generation of text are being used by climate denial groups to facilitate mass disinformation campaigns.”
“There is simply no evidence that AI will help the climate more than it will harm it,” said Jill McArdle, international corporate campaigner at Beyond Fossil Fuels.
“Rather than relying on credible and substantiated data, big tech companies are writing themselves a blank cheque to pollute on the empty promise of future salvation.”
Beyond Fossil Fuels was one of the six climate accountability groups involved in the report, which also included Stand.Earth, Climate Action Against Disinformation, Friends of the Earth, the Green Screen Coalition and the Green Web Foundation.
Additional reporting by Rei Takver
