Soaring infrastructure investments fuelled by AI innovation can accelerate the clean energy transition, according to a report which highlights stark differences between the paths being taken by firms such as Meta and Google.
The report, authored by US NGOs As You Sow and Sierra Club, warned that exaggerated forecasts of power demands by data centre were driving a rush to build new fossil gas plants by US utilities. But it stressed that tech companies have an opportunity to prove they can scale AI “while hitting climate goals, cutting costs for customers, and safeguarding shareholder value”.
The report was published ahead a wave of new investments in AI-related infrastructure by US firms in the UK, announced to coincide with this week’s state visit by President Donald Trump.
It contrasted the approach of tech conglomerate Google, which is planning to shift AI workloads between data centres to reduce electricity use during peak demand, thus easing strains on grid capacity, with that of social media giant Meta, which intends to power a new 2.2 gigawatt data centre campus via new methane gas plants.
US data centre electricity demand alone is projected to potentially triple by 2028, with estimates forecasting it to consume from 7% to 12% of total electricity consumption.
“Without transparency and smart planning, we risk building a fossil‑fuelled foundation that locks in high energy costs and climate damage for generations,” said Kelly Poole, As You Sow’s Climate and Energy Programme Coordinator. “Tech companies that embrace renewable procurement and demand flexibility will set themselves up to build faster and more reliably than those betting on gas and coal.”
Faced with “undiversifiable systemic risk” across portfolios from mounting and current climate costs, investors should pursue active engagement, the report said, by “pushing utilities to prioritise clean energy deployment and holding technology companies accountable for their energy use impacts”.
Among the new commitments announced by US firms, Microsoft said it would spend US$30 billion on AI-related development in the UK by 2028, with half going to cloud and infrastructure.
In a recent blog, Impax Asset Management said AI’s impact on energy and water resources can be managed, enabling the technology to support sustainability goals across industries.
“The digitalisation of processes, efficient cloud computing and advanced semiconductors can combine to radically reduce the resource intensity of the global economy.,” wrote Portfolio Manager Luciano Lilloy.
“We believe companies at the vanguard of innovation in these areas can present compelling long-term investment opportunities.”
