BERLIN: Environmental groups and non-governmental organisations (NGOs) are accusing major artificial intelligence companies behind the likes of Copilot, Gemini and ChatGPT of a "climate hoax" in glossing over the environmental impact of their services.
Tech companies like Microsoft, Google and OpenAI often justify the enormous energy appetite of their new data centres with the argument that AI is a crucial tool for tackling the climate crisis.
However in research published on Monday, several NGOs say that these claims rest on weak evidence. The study's authors accuse the industry of greenwashing and covering up the environmental damage they cause with misleading communications.
A central criticism in the study is the lack of differentiation in the use of the term artificial intelligence. The study shows that the positive climate effects promoted by companies such as Google and Microsoft relate almost exclusively to "conventional" AI applications – such as weather forecasting models.
The current boom, and the massive expansion of data centres that comes with it, is driven primarily by so-called "generative" AI for end users – systems such as ChatGPT, Copilot and Gemini that produce text, images and videos.
For these resource-intensive applications, the study's authors could find no example demonstrating a measurable and substantial reduction in greenhouse gas emissions.
The authors describe the linking of climate benefits from conventional AI with the expansion of generative models as a new form of greenwashing – a strategy of projecting a more climate-friendly image through misleading, vague or unsubstantiated claims about supposed environmental benefits, thereby distracting from the actual environmental damage caused.
For the analysis, NGOs including AlgorithmWatch and Beyond Fossil Fuels analysed 154 high-profile claims by technology companies and institutions about the positive climate effects of AI.
The results reveal a clear gap between promises and scientific evidence. Only 26% of the statements examined were based on published scientific studies. In 36% of cases, no evidence whatsoever was cited, while the majority of the remainder referred only to the companies' own websites or reports.
The authors conclude that even for conventional AI, the supposed climate benefits are often greatly overstated, while the negative effects of AI growth are clearly measurable.
Julian Bothe, senior policy manager at AlgorithmWatch, said that if there were sustainability benefits from artificial intelligence, they came from conventional AI applications with low resource consumption.
ChatGPT and other large language and image-generating models at the centre of the current AI hype consume vast amounts of electricity and water, produce CO2 emissions on a scale comparable to entire countries, and bring no positive environmental benefit whatsoever, he said. – dpa
