Artificial intelligence represents a rare historical window for developing nations to leapfrog stages of economic growth that took industrialised countries generations to navigate, according to a World Bank report released this week in London. The multilateral institution suggests that countries acting swiftly on three critical fronts—energy infrastructure, digital connectivity, and workforce capability—could harness AI's potential to deliver development gains at unprecedented speed. This assessment carries particular resonance for Southeast Asian economies still grappling with infrastructure deficits and labour market transitions, even as wealthier nations race to dominate AI innovation and deployment.
Indermit Gill, the World Bank's chief economist, framed the opportunity with striking urgency, describing AI as a "lifeline" that developing economies must seize before the window narrows. His comments reflect growing consensus among development economists that the AI revolution, unlike previous technological shifts, may offer pathways for lower-income nations to benefit without requiring the massive capital outlays or institutional frameworks that earlier breakthroughs demanded. The message challenges the widespread assumption that developing countries will inevitably lag in AI adoption and miss out on productivity gains that the technology promises.
The report's analysis reveals a counterintuitive finding about labour market vulnerability. Generative AI threatens employment in wealthy nations at triple the rate it does in poorer countries, with 14.2% of jobs at risk in high-income economies compared to just 4.5% in low- and middle-income countries. This disparity reflects structural differences in labour markets: advanced economies have concentrated financial services, professional roles, and administrative functions—precisely the tasks AI can automate most effectively. Developing nations, with larger populations engaged in agriculture, manufacturing, and informal services, face lower immediate displacement risk, though this advantage is temporary and conditional on strategic adaptation.
Yet the opportunities extend far beyond job preservation. Gill highlighted practical applications that could transform essential services across developing regions: health workers diagnosing diseases faster using AI-powered image recognition, educators designing personalised lesson plans, judicial systems processing cases more efficiently, and farmers making data-driven decisions about crop selection and planting timing. These applications require neither cutting-edge supercomputers nor proprietary large language models. Instead, modest AI tools adapted to local languages, climates, and institutional contexts could multiply the reach of limited professional expertise across sprawling populations with inadequate access to quality services.
The infrastructure challenge, however, remains formidable and demands honest acknowledgment. Global corporations are channelling billions into AI infrastructure—particularly the energy-intensive data centres that power training and deployment. Emerging economies cannot compete on this scale, nor do they need to for meaningful progress. But they do require reliable electricity supply, broadband connectivity, and an ecosystem of devices from smartphones to laptop computers. For many Southeast Asian nations, this means substantial continued investment in power generation, grid modernisation, and last-mile broadband expansion—priorities that compete for scarce government resources against healthcare, education, and poverty reduction.
Digital skills represent the third pillar, yet perhaps the most malleable challenge. Nations cannot wait for perfect infrastructure before building human capacity. Educational systems need to pivot toward computational thinking, data literacy, and AI ethics even as traditional literacy and numeracy gaps persist. This creates dual pressures: countries must simultaneously address foundational education shortfalls while preparing workforces for an AI-integrated economy. Malaysia and its neighbours have begun this transition, but scaling remains uneven across urban and rural areas, and between privileged and marginalised communities.
The International Monetary Fund's separate modelling suggests Sub-Saharan Africa's economy could expand by approximately 4% annually over the next decade if AI adoption proceeds favourably. While regional figures for Asia are less formally quoted, the magnitude of potential impact—whether positive or missed—is substantial enough to reshape growth trajectories and competitiveness within the next fifteen years. For Southeast Asia, where several nations aspire to upper-middle-income status, AI could either accelerate that transition or widen gaps with early-adopting peers.
Yet Gill's invocation of history carries implicit warning: developing economies that missed the first Industrial Revolution endured two centuries of relative decline and dependency. The analogy, while striking, may overstate AI's uniqueness—technological adoption is rarely binary, and catching-up opportunities often persist longer than doomsayers predict. Nonetheless, the underlying point resonates: early-movers gain strategic advantage, institutional learning, and positioning within global value chains that late-comers struggle to match.
The World Bank flagged substantial risks that complicate the optimistic narrative. AI could intensify income inequality if benefits concentrate among educated urban workers while displacing others; misinformation campaigns could become more sophisticated and persuasive as generative tools improve; and authoritarian governments might weaponise surveillance and profiling capabilities embedded in AI systems. These dangers are not hypothetical—they reflect demonstrated capacities of current tools and documented government practices across several developing regions. Policymakers cannot simply adopt AI wholesale; they must weave in safeguards, accountability mechanisms, and regulations designed for local contexts.
For Southeast Asian policymakers and business leaders, the World Bank's assessment suggests a middle path between techno-optimism and resignation. AI is neither a magic solution that solves development challenges automatically nor an existential threat to be feared and blocked. Instead, it is a toolkit—powerful and flexible—that yields different outcomes depending on how governments, educators, businesses, and communities integrate it into existing institutions and priorities. Success requires simultaneous progress on electricity, connectivity, skills, regulation, and inclusion, not sequential achievement of perfect conditions.
The narrowing window is real but not absolute. Unlike earlier technological revolutions where first-movers locked in decades-long advantages, AI's relatively open source ecosystem, falling compute costs, and adaptability to local problems create space for deliberate catch-up strategies. Governments that articulate clear AI adoption policies, invest in complementary infrastructure, and protect vulnerable populations stand to convert the "lifeline" into sustained development gains. Those that drift without strategy risk finding themselves further from the boat when the current shifts.
