The UNDP report "The Energy Sprint of the AI Race: A Green Window of Opportunity for Developing Countries?" examines the growing relationship between artificial intelligence, energy systems, and sustainable development. It argues that access to abundant, reliable, affordable, and clean electricity will increasingly determine countries’ ability to participate in the AI economy. For developing countries, renewable energy could create a “green window of opportunity,” but only if it is combined with stronger infrastructure, skills, institutions, investment, and policies that ensure local economic and social benefits.
Key Insights
AI is becoming increasingly energy-intensive, making reliable and affordable clean electricity a major factor in future AI competitiveness.
Global data-centre electricity consumption is projected to more than double by 2030, while demand from AI-focused data centres could triple.
AI could add around US$15.7 trillion to global output by 2030, but only about 10% is expected to benefit developing economies.
Countries with strong renewable energy resources could gain new opportunities in green data centres, AI-enabled energy systems, climate applications, and digital services.
Renewable energy potential alone is not enough; countries also need reliable grids, digital infrastructure, skilled workers, effective regulation, financing, and local demand.
The report promotes “Green AI,” which combines clean energy, responsible resource use, environmental protection, and human development.
Developing countries risk remaining suppliers of land, energy, minerals, and data while higher-value AI activities and profits are captured elsewhere.
The report proposes a people-planet-prosperity framework that links inclusion and skills, environmental sustainability, and economic transformation.
Five priorities are recommended: strengthen local AI-energy ecosystems, make Green AI the default approach, move up the minerals-energy-compute value chain, align AI with energy and industrial strategies, and support Green AI startups.
For many developing countries, a phased approach may be more realistic: initially investing in AI applications, skills, and institutions before committing heavily to expensive large-scale computing infrastructure.