The World Bank released the assessment on Tuesday, describing it as the first comprehensive look at how generative AI could affect the 6.8 billion people living in developing economies. The report portrays a technology with uneven consequences: lower average job exposure than wealthy nations face, productivity gains that currently favor the rich, concentrated losses in a handful of industries, and the risk of new dependence on the United States and China.
The job-disruption gap
The World Bank calculated that generative AI threatens 4.5% of jobs in developing countries, compared with 14.2% of jobs in rich ones. The gap reflects the structure of poorer economies, the bank said. “Still agrarian and reliant on small enterprises, poorer economies are less likely to suffer large-scale job losses: AI is more likely to lend their workers a hand than put them out of work,” the bank said.
“As a rule, developing economies today have more to gain—and less to fear—from AI than richer ones,” wrote Indermit Gill, the World Bank’s chief economist.
Where the upside depends
The bank built its case for the upside on what it described as a long-standing obstacle to growth in developing countries. “Many developing countries face shortages of highly skilled workers,” the bank said. “AI can help by enabling less experienced workers to perform more advanced cognitive tasks, allowing some work to be shifted away from scarce specialists and making existing workers more productive.”
Those gains do not require the heavy infrastructure now reshaping the U.S. economy, where the race to develop frontier AI has drawn hundreds of billions of dollars into data-center construction, the bank said. “No need for massive investments in data centers or large language models,” Gill wrote. “AI can help solve important problems involving narrow tasks even when local computing power is limited, electricity is unreliable, and internet service is dodgy.” Many developing economies lack the ability to generate the vast amounts of electricity that such data centers would demand, the bank said.
For now, however, the bank said rich countries stand to benefit most from the spread of AI. If current adoption trends hold, the bank calculates that productivity gains will be more than three times larger in rich economies than in developing ones. The bank said that gap could be closed if developing economies expand access to computing power and improve the availability of local data, including in local languages. “AI needs to be customized to the local context because it learns from the data upon which it is trained,” the bank said.
Losses land unevenly
Job losses will not be spread evenly. The bank flagged particular risk for developing economies that have relied on the outsourcing of business processes, such as call centers and back-office services. “AI could close off a promising route to middle-class employment in many developing economies, threatening call-center work and entry-level jobs in software, finance, and business services,” Gill wrote.
The dependence trade-off
The bank also warned that embracing the technology could leave developing countries dependent on two economic giants. “The world’s most advanced AI systems are controlled by a small number of companies, mainly in the United States and China,” the bank said. “This means developing countries could become dependent on technology they do not control.” That threat could deter leaders from backing the widespread deployment of AI at the risk of missing out on its economic benefits, the bank said.
Rather than spending the large sums required to replicate the AI supply chain, the bank offered what it called a more practical route to limiting that exposure. “A more practical way to reduce dependence on a single country is to buy models, cloud services, and other AI tools from many countries—and make sure they can work together and be swapped out without requiring the entire system to be rebuilt,” Gill wrote.