🎙 Listen to a summary of this story
Asia’s race to embrace artificial intelligence is colliding with another urgent challenge: cutting the emissions that drive climate change. The technology promises better weather warnings, more efficient factories and smarter electricity networks. But powering it requires expanding an energy system already struggling to meet rising demand.
That tension is reshaping the region’s economic opportunities. Countries producing chips, servers and other AI equipment are benefiting from investment and exports, while access to reliable, affordable electricity is becoming increasingly important to their prospects.
The divide is visible in the World Bank’s October 2026 East Asia and Pacific Economic Update. It finds that AI-related manufacturing is lifting growth in participating economies, while high energy costs weigh on others, particularly Pacific island nations.
Yet the evidence does not support a simple division between countries attracting money through AI and green credentials and those being left behind. Climate technology funding has faced its own slowdown, and an expanding digital economy does not automatically become a cleaner one.
AI’s growing appetite for electricity
The scale of the energy challenge is becoming clearer. The International Energy Agency’s 2026 assessment projects global data center electricity consumption will roughly double, from 485 terawatt-hours in 2025 to 950 terawatt-hours in 2030 — around 3% of worldwide electricity demand.
Those figures cover all data centers, including facilities supporting conventional digital services. Electricity consumption at AI-focused centers is projected to triple over the same period.
The pressure is especially concentrated around major computing hubs. In its 2025 regional analysis, the agency projected Southeast Asian data center electricity demand would more than double by 2030, partly reflecting development around Singapore and southern Malaysia.
Singapore has explicitly linked expansion to sustainability. Its Green Data Centre Roadmap, announced in 2024, targeted at least 300 megawatts of additional capacity in the near term, alongside efforts to improve efficiency and expand green energy use.
The broader climate implications depend on how additional electricity is generated. Renewable power can limit emissions, while reliance on fossil fuels adds to them. Meeting a company’s annual electricity use with renewable purchases also differs from supplying clean power whenever its servers operate.
Efficiency gains offer some relief. The IEA says energy consumption per AI task has fallen sharply. But increasingly popular applications, including video generation and complex reasoning, can consume far more electricity than simple text responses, complicating forecasts.
From smarter grids to earlier warnings
AI can also help energy systems operate more efficiently. Forecasting tools can anticipate changes in wind and solar production, while monitoring systems can identify equipment problems before they cause outages.
An IEA analysis of energy applications estimated that AI could help unlock as much as 175 gigawatts of additional transmission capacity on existing power lines. That represents potential from wider deployment, rather than capacity already delivered.
Such tools could help countries connect renewable generation more quickly. They still require reliable measurements, trained operators and investment in physical equipment. Software cannot substitute for every new transmission line or transformer.
Weather forecasting offers another practical application. The World Meteorological Organization launched an Asian AI nowcasting pilot in Shenzhen in May 2025, bringing together meteorologists and technology specialists to evaluate systems forecasting conditions over the next few hours.
The initiative includes preparations for deploying tools in developing countries. Its focus on testing and comparing forecasts is important: faster predictions must also be dependable enough to support decisions about dangerous rainfall, lightning and other hazards.
For communities exposed to extreme weather, better forecasts can support earlier action. But their usefulness depends on warnings reaching people and authorities having the resources to respond. These systems support adaptation to climate risks; they do not themselves prevent warming.
Investment follows practical returns
The funding picture is less straightforward than a seamless convergence of AI and climate investment. In a December 2025 interview published by ADB Ventures, senior fund manager Jugnu Pati described weakening climate technology funding, particularly in Asia, while identifying opportunities for commercially useful solutions.
Her examples included businesses applying AI to fertilizer use, vehicle performance and renewable energy. Their appeal rests on solving operational problems and demonstrating savings, rather than simply combining two fashionable investment themes.
The World Bank similarly argues that most regional economies can gain in the short term by adopting and adapting existing AI tools, especially smaller applications. Building the largest models is not a prerequisite for benefiting.
The emerging policy test is therefore practical: whether AI delivers measurable benefits while its infrastructure receives scrutiny over energy use and emissions. For Asia, attracting computing investment and tackling climate change can reinforce each other, but neither outcome is guaranteed.
