‘A short-term hit for long-term benefit’: How these ESG managers justify investing in AI

Managers from Rathbones and RLAM explain how they navigate the energy challenges of investing in tech

Sustainable Artificial Intelligence AI Energy, renewable energy source, battery technology, green energy
3–5m

Artificial intelligence (AI) has delivered incredible returns, with the MSCI ACWI Information Technology index up 123.9% over the past five years (as of 20 April 2026), but investing in these businesses can be at odds with the goal of sustainable funds.

AI data centres demand enormous energy to run and set up, with research from the International Energy Agency finding energy consumption from AI data centres will double by 2030, with power use set to triple in the same period.

While this may seem incompatible with the goals of ESG-focused funds, a handful of sustainable investors have argued there are still opportunities for AI stocks in ‘green’ portfolios.

This is the argument of David Harrison, manager of the Rathbone Greenbank Global Sustainability fund, which currently holds Microsoft and Nvidia as its two largest stocks. Leading semiconductor chip manufacturer TSMC also features in its top 10 holdings.

In total, Harrison rated his tech and power exposure in the portfolio as roughly 25%.

This level of exposure does pose a challenge as a sustainable investor, he conceded. Data centres have “truly staggering” energy demands, with a single data centre in parts of the US requiring as much as a gigawatt of power to run, he said.

That said, “the opportunity set in AI for sustainable investing is still vast”, he told Portfolio Adviser.

“There’s a balance to be struck in this,” he said. “You can see the benefit of this spend in multiple industries, but companies need to articulate and disclose what their energy demand is before we can invest in them.”

See also: Five themes for sustainable investment in 2026

This is easier to do with some companies than others, with companies further up the supply chain with big spending on AI capex, such as Microsoft, disclosing this more readily than smaller companies, he said.

He is not the only sustainable investor to argue AI companies are not incompatible with sustainable goals.

Mike Fox, head of sustainable investments and equities at Royal London Asset Management, noted: “We accept the point on energy intensity and the social consequences that can come from AI.”

However, he argued this might be more of a “short-term hit in exchange for a long-term benefit”.

“The key issue as a sustainable investor is if you think AI is a productivity benefit to the broader economy, somewhere down the line you should see energy efficiency from it,” the RLAM manager explained.

He pointed to Google, which bought AI company DeepMind and rolled out its technologies across the wider business, leading to a “considerable” reduction in energy usage.

“We’re mindful of the energy challenges, but we think in the end that can be worked through,” Fox concluded

See also: Sustainable fund outflows slow in Q4

Opportunities for sustainable investors

One way to play on AI without some of the energy concerns is through the “picks and shovels” of grid infrastructure, according to Rathbones’ Harrison.

He pointed to Belimo, a Swedish company that makes actuators for air-conditioning units that are used to regulate the temperature in commercial businesses and data centres.

“When you’re running an air conditioning unit, you need a company like Belimo there to make it more efficient,” the Rathbones manager said.

For Harrison, this is a great example of a stock that can benefit from the rise in AI, while maintaining sustainable goals.

“If you think about it from that sustainability angle, those companies actively making data centres more energy efficient and providing a solution to that power output question, it’s a name we really like,” the Rathbones manager said.

He added the issue of AI engagement was perhaps “the biggest topic for [Rathbones]” so far this year, with virtually all companies they own, or are interested in, examining how to apply AI to their businesses.

This was occurring even in businesses that would initially seem unrelated to advances in AI, such as consumer goods companies or medicine, he noted.

“You hear a lot about AI in terms of how sticky it is, but you can see so many examples from companies about how they’re using it.”

He highlighted medical companies such as Boston Scientific, a leading manufacturer of heart valve technology.

See also: Keeping the lights on: How utilities are meeting AI’s energy needs

“They spoke very coherently about developing new products incorporating AI, such as CAT scans or medical imaging,” he explained.

More generally, he said AI is being used for medical advancements in these companies, which is another way to expose the portfolio to AI without some of the associated ESG worries.

“We recognise AI is a complex issue as a sustainable investor, but I really believe there’s a lot of opportunity out there, it’s just about how companies are engaging with it,” Harrison concluded.