A Counterintuitive Finding
In its landmark World Development Report 2026, the World Bank presents a picture that defies common assumptions about AI's global impact. The report finds that jobs in high-income countries are more than three times as likely to be at risk from generative
AI automation than those in developing economies. Specifically, it estimates that 14.2% of jobs in wealthy nations are exposed to automation, compared to just 4.5% in low- and middle-income countries. This data suggests that the immediate threat of AI-driven job displacement is less severe in nations like India than in the United States or Europe. The primary reason for this disparity is the fundamental structure of these economies. Rich countries have a higher concentration of knowledge-based, white-collar sectors such as finance, marketing, and tech—jobs heavy on cognitive tasks that current AI models are well-suited to perform. In contrast, many developing economies are still anchored in sectors less susceptible to today's AI.
The Structure of Work Matters
The lower exposure in developing nations stems from the nature of their labour markets. A significant portion of the workforce is engaged in agriculture, manual labour, and small-scale enterprises where tasks are less routine and codifiable, making them harder and less economical to automate. According to the World Bank, AI is more likely to augment these workers' capabilities rather than replace them outright. For example, AI can provide farmers with precise weather forecasts, help community health workers with medical screenings, or assist teachers in creating lesson plans. The report highlights that the biggest gains from AI in these countries will come from amplifying human capabilities to fill expertise gaps, such as the scarcity of doctors or agricultural experts. In this view, AI serves as a powerful tool for productivity, with 16.2% of jobs in developing economies poised for a meaningful boost, a figure close to the 18.7% expected in high-income countries.
A Double-Edged Sword
However, this lower automation risk is not purely good news. It is also a symptom of underlying structural challenges. Many developing countries lack the foundational infrastructure—reliable electricity, widespread internet access, and digital skills—necessary to adopt AI at scale. As of 2024, nearly a third of rural schools in Sub-Saharan Africa lacked dependable power, and over two-thirds lacked internet. This digital divide means that while fewer jobs are currently at risk, these economies are also in danger of being left behind as the world shifts towards an AI-driven paradigm. The World Bank warns that the window of opportunity to catch up is narrow. If developing nations fail to invest in these fundamentals, they risk widening the productivity gap with advanced economies, potentially leading to what some officials have termed a 'lost decade' of growth.
The Indian Context and Key Risks
For India, the report's findings are particularly relevant. While some World Bank data has previously suggested a high long-term automation risk for up to 69% of jobs, the recent 2026 report paints a more nuanced, immediate picture. The country's large agrarian and informal sectors provide a temporary buffer against mass displacement. However, the report issues a critical warning for sectors that have been a key engine of middle-class employment: business process outsourcing (BPO), call centres, and entry-level IT services. These roles are highly susceptible to automation by AI, which could close off a traditional pathway to economic mobility for many. This highlights a key challenge for India: balancing the protection of vulnerable sectors while fostering innovation and adaptation in its globally-integrated services industry.
The Path Forward: Adopt, Adapt, Advance
The World Bank urges developing countries not to sit idle. It proposes a clear three-step framework: Adopt, Adapt, and Advance. The first step is to adopt existing low-cost AI tools to improve public services and business productivity. Critically, these tools must then be adapted to local contexts, languages, and data to be effective. Simply importing a model built for a high-income country will not work. Finally, over the long term, nations can aim to advance their own AI capabilities. This strategy doesn't require building massive, costly AI models from scratch. Instead, the focus is on practical application and solving local problems. Governments are encouraged to use their purchasing power to promote AI solutions in key areas like healthcare and education and establish the foundational skills and infrastructure needed for widespread use.
