Google has revealed its emissions have increased by 48% over the last five years due to the increasing energy demand from AI.
The AI energy demand is also making it more difficult for the company to meet its net zero 2030 goals, according to a new Google report.
The annual report was released yesterday and covers Google’s progress toward meeting its environmental goals last year.
The Alphabet Inc. unit said its greenhouse gas emissions totalled 14.3 million metric tonnes of carbon dioxide equivalent throughout 2023.
This is 48% higher than in 2019 and 13% higher than in 2022.
Why is AI energy demand so high?
AI — and in particular generative AI, which takes in user inputs and spits out new content like text, images or songs — is extremely resource-intensive
As the technology grows rapidly, more and more data centres are needed to build and run it, leading to surging energy demand.
Google commented: “AI has been deeply integrated into our products for years, and we’ve invested heavily in improving the efficiency of our AI models and infrastructure.
“While these efforts have helped mitigate some of AI’s environmental footprint, the rapid advancement of AI has brought necessary increased attention to its energy consumption and resource demands.”
The report says that predicting the future environmental impact of AI energy demand is complex and evolving.
The role of data centres in increasing emissions
Google blamed higher energy consumption at its data centres and emissions from its supply chain, and its push to add AI to its products could make it more difficult to reduce emissions going forward.
According to the International Energy Agency, global data centre electricity consumption is 240–340 TWh or around 1–1.3% of global electricity demand.
Using these estimates as a proxy for 2023, Google’s data centre consumption of more than 24 TWh in 2023 translates to approximately 7–10% of global data centre electricity consumption and approximately 0.1% of global electricity demand.
However, while data centres consume electricity and contribute to emissions, cloud and hyperscale data centres collectively represent only an estimated 0.1–0.2% of global GHG emissions.
The report stated: “Google’s emissions contribute a smaller fraction—in 2023, our total GHG emissions were approximately 14.3 million tCO2 e.”
Uncertainty ahead
AI is at an inflection point, and many factors will influence its ultimate impact on the environment—including the extent of AI adoption, the ability to mitigate its footprint, and the pace of continued innovation and efficiency.
The report concludes that system-level changes are needed to address challenges such as grid decarbonisation, evolving regulations, hard-to-decarbonise industries, and the availability of carbon-free energy.
It states: “While we remain optimistic about AI’s potential to drive positive change, we’re also clear-eyed about its potential environmental impact and the collaborative effort required to navigate this evolving landscape.”
“We’re committed to responsibly managing the environmental impact of AI by deploying three major strategies: model optimisation, efficient infrastructure, and emissions reductions.”
