3 min read

Tech giants urge transparency from data centre vendors amid AI-driven emissions rise

Embodied emissions from manufacturing and construction are estimated to make up 6.6% of data centres' Scope 3 footprint.
Melodie Michel
Tech giants urge transparency from data centre vendors amid AI-driven emissions rise
Photo by Jordan Harrison on Unsplash

Tech behemoths including AWS, Google, Meta and Microsoft have signed an open letter urging data centre suppliers to disclose their embodied emissions as they struggle to contain the growing carbon footprint of their AI applications.

The Big Tech firms (all of which recently reported significant emissions rises as a result of data centre expansion to power their AI activities), along with Digital Realty and Schneider Electric, argue in a letter published this week that their ability to mitigate these emissions is hampered by the lack of availability of Environmental Product Declarations (EPDs) from vendors.

This lack of transparency not only prevents them from procuring low-carbon materials, but also hinders their own climate reporting, “which can jeopardise contracts with some customer groups”. 

As a result, the six companies, which form the governing body of the iMasons Climate Accord, launched in 2022 to develop a standard for the digital industry’s carbon reporting, are urging their data centre equipment and materials suppliers to create EPDs aligned with existing standards for their products, and to make them readily available.

“By taking these actions, our supply chain vendors will improve the transparency of digital infrastructure embodied emissions, support our GHG estimation and reporting, and help ensure we are positioned to meet our climate goals via more informed green procurement as we develop the digital infrastructure of the future,” they write.

Big Tech in a sustainability pickle over AI expansion

While artificial intelligence is touted as a game-changer for corporate sustainability (Microsoft’s Chief Sustainability Officer Melanie Nakagawa recently said generative AI freed up no less than 70% of her team’s time), its negative climate impacts are now becoming clear.

Microsoft itself reported a 30% emissions rise in 2023 alone, due to the construction of new data centres to meet growing cloud and AI expectations. Google’s emissions grew by 13% the same year, and 48% since 2019, prompting CSO Kate Brandt to warn stakeholders that achieving Google’s target of net zero emissions by 2030 will require navigating a lot of uncertainty, including “around the future of AI’s environmental impacts”.

Digital infrastructure experts have long known that power-hungry AI is a potential ticking time bomb, made worse by a lack of transparency from cloud vendors around their data centre emissions. 

In the letter, the companies recognise the digital industry’s responsibility to mitigate climate change by lowering the greenhouse gas (GHG) footprint of its data centres, and says it is already setting targets and deploying solutions. 

(For Microsoft, these solutions include investing massively in carbon capture and storage.)

How much of a data centre’s carbon footprint is embodied emissions?

Much of the concern over AI’s environmental impacts is around data centres’ power usage. The International Energy Agency (IEA) estimates that data centres, cryptocurrencies, and artificial intelligence (AI) consumed about 460 TWh of electricity worldwide in 2022, almost 2% of total global electricity demand – but their power demands could double by 2026.

As of 2023, data centres and data transmission networks were responsible for 1% of the world’s energy-related emissions, but despite Big Tech’s large purchases of renewable electricity, recent emissions data suggests this number is bound to increase in the coming years.

But beyond operational emissions from data centres’ power usage, little is known about data centres’ lifecycle carbon footprint. A pilot study published by Schneider Electric last year estimated that Scope 3 could represent between 38% and 69% of a data centre’s carbon footprint, and that embodied emissions from manufacturing and construction make up 6.6% of that Scope 3 footprint.

The UN Environmental Programme (UNEP) recently urged stakeholders to consider all the environmental implications of AI’s expansion – “such as increased demand for critical minerals and rare earth elements and water resources to meet data centre demands” in a report warning against “short-termism” in technological development.