The environmental and financial implications of data centres in the context of further AI development is explored in a new report by the Environmental Defense Fund and Nuveen, entitled Decoding data centers: Sustainability due diligence across the value chain.
As the report notes, AI is an ‘infrastructure story playing out across many sectors, geographies, and asset classes’, which in turn is creating both opportunities and environmental risks for businesses.
‘Investor attention’
“Artificial intelligence is gripping investor attention as technology advances reshape global markets with high uncertainty about the industry’s path,” commented Sarah Wilson, managing director, head of Climate Center of Excellence, Nuveen. “What has emerged already is a better understanding of the environmental externalities and implications for climate change of data centres’ rapidly increasing energy use.”
Three main environmental risk factors are explored in the report, with energy use the primary area of concern, as data centres continue to drive ‘soaring demand’ for electricity generation.
Water usage is identified as a further issue, given the impact that data centres can have on water stress and water quality. A third area of concern regards the impact on communities, with the construction and operation of data centres leading to increased truck traffic, noise, air pollution, land-use conflicts, and higher utility expenses.
Among the key insights discussed in the report is the fact that only a handful of technology firms play ‘crucial roles’ in the data centre ecosystem, however many players have responsibility for mitigating environmental risks.
In addition, the report notes that efficiency is a key aspect of managing energy demand across the AI value chain, particularly when it comes to chip manufacturing. It also calls on all stakeholders, including investors, to conduct appropriate due diligence on environmental and social impacts to manage the financial risks associated with data centre operations.
‘Competing effects’
Wilson added that the energy transition and decarbonisation of data centres faces “two competing effects” – increased energy and grid demand may encourage greater fossil fuel use, delay plant retirements, and drive new gas infrastructure, while at the same time, growth in clean energy deployment, procurement, and investment supports accelerated commercialisation of carbon-free technologies.
“Increasingly for hyperscalers, the ability to power compute is a competitive differentiator and scaling strategies will influence the balance of these two opposing forces,” she said.
“For investors, there are tremendous opportunities, and risks — both idiosyncratic and systemic — in the AI boom. We encourage our peers and the investor ecosystem writ large to utilise this guide for environmental stewardship and risk mitigation and to elevate market expectations for investees throughout the diverse value chain.” Read more here and here.
