Subscribe
Jobs in high-income countries are more than three times as likely to face automation risk from generative AI as those in low- and middle-income countries, according to a World Bank report.
The report says AI could boost productivity in more than 16% of existing jobs in developing economies, compared with more than 18% in advanced economies.
It warns that developing countries still lack many of the basics needed to use AI effectively, including reliable electricity, internet access, data systems, skills and institutions.
Separate research on agriculture found that AI could improve productivity and resilience, but only with supportive policies, infrastructure and equitable access.
Jobs in high-income countries are more than three times as likely to be exposed to automation by generative artificial intelligence (AI) as those in low- and middle-income countries, according to a new World Bank report.
The World Development Report 2026: The Promise of Artificial Intelligence found that 14.2 per cent of jobs in high-income countries face potential automation risk from generative AI, compared with 4.5 per cent in low- and middle-income countries.
But the report also said AI could boost productivity in more than 16 per cent of existing jobs in developing economies, compared with more than 18 per cent in advanced economies. “The greatest promise for developing countries lies not in replacing workers, but in amplifying what they can do,” the report said.
The findings come as developing economies face a more uncertain global outlook. A May 2026 report by the United Nations Trade and Development warned that the global economy was moving beyond supply disruptions and inflation into a more fragile and prolonged period of uncertainty, with developing countries particularly vulnerable to financial stress and shortages.
The report comes amid growing debate over the ethical use of AI and concerns that it could make sections of the workforce redundant. But the World Bank said AI is already helping people, businesses and governments solve problems, analyse information, improve forecasts and deliver services at a larger scale.
These tools could be especially useful in developing countries, where trained professionals, reliable records and public capacity are often limited, the report said.
AI could help doctors diagnose patients, support farmers in making crop decisions, help businesses improve productivity and assist governments in areas such as tax collection and disaster response.
A December 2025 study comparing AI adoption in agriculture across developed and developing countries reached a similar conclusion. It examined evidence from technologically advanced economies, including Europe, the US, Australia and Japan, alongside studies from Africa, South Asia, Latin America and other low- and middle-income regions.
The study found that AI has strong potential to improve agricultural productivity and resilience. But it said those gains depend on supportive policies, reliable infrastructure and equitable access. Without them, AI could reinforce existing inequalities rather than reduce them.
Indermit Gill, chief economist at the World Bank Group, said developing economies may no longer need “large models or big data centres” to benefit from AI. “But they must hurry: AI is spreading faster and is more context-specific than earlier general-purpose technologies like electricity and the internet,” he warned.
The report said AI could offer tangible benefits at a time when developing countries are facing their weakest growth average in more than three decades.
The opportunities, however, are not guaranteed. “The window to get this right is narrow,” said Gaurav Nayyar, director of the World Development Report 2026. “AI presents a once-in-a-lifetime opportunity to solve problems that have resisted solutions for generations.”
The report warned that most advanced AI models are being built by a small number of countries and companies. By contrast, many developing countries still lack the basic conditions needed to use AI effectively, including reliable electricity, internet access, data systems, skills and strong institutions.
Without policy correction, AI could further widen the inequality gap between the countries and increase newer “risks for safety, rights and social cohesion”, the report warned.
Hence, the three step approach of adopting new tools and customising them to local conditions and finally moving onto “frontier” or advanced AI models, it argued. But on ground, the picture is stark: in sub Saharan Africa, almost one-third of schools lack electricity while another two-third lacks internet access, the report highlighted.
