AI boom triggers increase in ESG risk consideration

The uplift in artificial intelligence (AI) adoption has also correlated to a spike in consideration of the megatrend’s environmental, social and governance (ESG)-related risks, according to a new report from Capital Group.

The Capital Group ESG Global Study 2025, based on research collected from 1,130 global investors across 25 regions, found while global ESG adoption had slightly slowed (87 per cent) compared to 2024 and 2023 (both 90 per cent), there had been a sharp upswing in the number of investors identifying AI-related emissions and effects as their top ESG risks.

Approximately 73 per cent of investors identified energy consumption and greenhouse gas emission as the top risks in this year’s report, compared to just 54 per cent last year. More than 40 per cent said the level of water consumption was also concerning, compared to just 18 per cent in 2024.

Concerns over AI’s “energy-intensive nature” presenting as a threat to progress made on the transition to net zero were also identified by the report, with 58 per cent of respondents agreeing. However, 56 per cent of respondents agreed that AI will play a part in “accelerating innovation in decarbonisation-related technologies and products, which will in turn help the energy transition” and 45 per cent agree that AI will help “improve energy
efficiency and grid operation, which will offset the impact from its high electricity usage”.

“This year’s ESG Global Study highlights the enduring role of ESG in the investment process as investors continue to evolve their approach. The consideration of ESG issues in fixed income and private markets are also gaining traction,” Jessica Ground, Global Head of ESG at Capital Group, said.

“As the AI boom continues, investors are more alert than ever to the environmental risks posed by AI – especially around energy use and water consumption. At the same time, they’re exploring how AI can drive innovation and accelerate progress on energy transition.”

Approximately 50 per cent of respondents also said their organisation – when it invests in companies that derive revenue from AI – will preference the more environmentally-sustainable companies over those with “poor environmental practices”.