AI Drives 48% Increase in Google Emissions

Google’s greenhouse gas emissions surged by 48% in 2023 compared to 2019, according to the tech giant’s latest environmental report. The spike is largely attributed to the increasing energy demands of data centres, driven by the rapid growth of artificial intelligence (AI).

AI-powered services require significantly more computing power—and consequently more electricity—than traditional online activities, raising concerns about their environmental impact. Google’s goal is to achieve net zero emissions by 2030, but the company acknowledges that integrating AI into its products will make this target challenging.

In its 2024 Environmental Report, Google cites the “greater intensity of AI compute” as the main factor for the heightened energy demands. Data centers, which house vast numbers of computer servers, are particularly energy-intensive when running AI systems. For instance, generative AI systems like ChatGPT can consume up to 33 times more energy than task-specific software, a recent study suggests.

Google’s report also highlights global disparities in the environmental impact of its data centers. Facilities in Europe and the Americas largely rely on carbon-free energy sources, while those in the Middle East, Asia, and Australia use significantly less carbon-free energy. Overall, about two-thirds of Google’s energy comes from carbon-free sources.

Tom Jackson, a professor of information and knowledge management at Loughborough University, emphasizes the often-overlooked environmental footprint of data storage. “People don’t realize everything they’re storing in the cloud is having an impact on their digital carbon footprint,” he says. Jackson leads the Digital Decarbonisation Design Group, which aims to measure and reduce the carbon footprint of data usage.

One major contributor to this footprint is “dark data”—data collected by organizations but rarely or never used. Despite being inactive, storing dark data on chips consumes substantial energy. “On average, 65% of the data an organization stores is dark data,” Jackson notes. He commends Google’s net zero goal for its data centers by 2030 but acknowledges it will be “really tough” to achieve.

The growing energy and water consumption of AI has sparked several warnings, especially given the sector’s projected rapid expansion. In March, the head of the UK’s National Grid predicted a six-fold increase in energy demand over the next decade due to AI and quantum computing.

However, some tech leaders, like Microsoft co-founder Bill Gates, downplay the environmental impact of AI. Speaking in London last week, Gates estimated AI would boost electricity demand by only 2% to 6%, but he believes AI advancements could offset this increase by driving significant energy savings in other areas. “The question is, will AI accelerate a more than 6 per cent reduction? And the answer is: certainly,” Gates said, as reported by the Financial Times.