Google’s greenhouse gas emissions in 2023 were 48% higher than in 2019, according to its latest environmental report. The tech giant attributes this surge to the increasing energy demands of its data centres, driven by the explosive growth of artificial intelligence (AI). AI-powered services require significantly more computing power—and consequently more electricity—than standard online activities, raising concerns about the technology’s environmental impact.
Google’s target is to reach net zero emissions by 2030, but the company acknowledges that “as we further integrate AI into our products, reducing emissions may be challenging.” In its 2024 Environmental Report, Google cites “increasing energy demands from the greater intensity of AI compute” as a key factor. Data centres, essentially massive collections of computer servers, are critical to AI operations, requiring enormous energy consumption.
A generative AI system, such as ChatGPT, might use around 33 times more energy than machines running task-specific software, according to a recent study. This highlights the substantial energy footprint of advanced AI technologies.
However, Google’s report also reveals significant global disparities in the impacts of its data centres. Most centres in Europe and the Americas derive the majority of their energy from carbon-free sources, in stark contrast to those in the Middle East, Asia, and Australia, which rely more heavily on non-renewable energy. Overall, about two-thirds of Google’s energy is sourced from carbon-free options.
“If you actually go into a data centre, it’s really hot and really noisy,” says Tom Jackson, professor of information and knowledge management at Loughborough University. “People don’t realise everything they’re storing in the cloud is having an impact on their digital carbon footprint,” he adds. Professor Jackson heads the Digital Decarbonisation Design Group, which aims to measure and mitigate the carbon footprint of data usage.
“Data providers must work closely with large organisations to help them reduce their dark data storage,” he explains. Dark data refers to information collected by organisations that is either used infrequently or not at all. Despite its lack of use, storing dark data on chips still consumes substantial energy.
“On average, 65% of the data an organisation stores is dark data,” says Professor Jackson. He commends Google’s target of achieving net zero emissions in its data centres by 2030 but acknowledges that it will be “really tough.”
The rising energy—and water—consumption of AI has sparked numerous warnings, particularly as the sector is expected to continue its rapid growth. The head of the UK’s National Grid predicted in March that AI and quantum computing would lead to a six-fold increase in energy demand over the next decade.
For more information, visit BBC News.