AI’s impact on jobs may be smaller in developing countries

Artificial intelligence is transforming the global workforce, but its impact may not affect all regions equally. Much of the conversation about AI and jobs focuses on high-income countries—where the technology threatens to reshape entire industries. But what will AI mean for workers in developing nations, who constitute 80 percent of the global workforce?

To better understand AI’s labor market impact in the developing world, in a recent paper we analyzed data from 25 countries, covering a population of 3.5 billion people. For workers in those countries, we assessed the level of AI exposure, which captures to what extent their jobs could be performed using AI. Our findings suggest that AI’s effects on jobs will be more gradual in the Global South, particularly in low-income countries.

AI exposure differs by income level

There is considerable variation in the exposure of jobs to AI: some occupations, like roofers, have very low exposure, meaning AI is unlikely to affect their tasks. At the other extreme, jobs like payroll clerks face high exposure, indicating that AI could significantly change how they operate. Globally, the greatest concentration of workers is in occupations with medium-low levels of AI exposure, such as motor vehicle mechanics.

Distribution of AI Occupation Exposure for Workers in All Countries

Source: Authors’ analysis.

A key takeaway from our analysis is that workers in low-income countries experience significantly lower AI exposure than those in high-income countries, with middle-income countries falling in between. This is partly due to a different labor market structure in developing countries with more jobs involving manual labor or interpersonal interaction. These types of jobs are less amenable to the changes brought about by AI. We also show that lack of access to electricity and internet further limits exposure particularly in low-income countries.

Source: Authors’ analysis

Our research also highlights demographic differences: women tend to have higher AI exposure than men—but only in high and upper-middle-income countries. Meanwhile, unlike in wealthier nations, where older workers are the most exposed to AI disruption, age-related differences in AI exposure are minimal in developing economies.

What is AI exposure?

To calculate our measure of AI exposure, we start with the set of tasks which make up each job. Next, drawing from previous work, we assign an AI exposure level to each task. Then we aggregate to get the AI exposure value for each job. Finally, we map the AI exposure by job to each individual worker in labor force survey data from around the globe.

It is important to clarify that AI exposure does not necessarily mean that a task—or an entire job—will be replaced. Exposure could mean three things:

1. Automation – AI fully takes over certain tasks, reducing the need for human labor.
2. Augmentation – AI enhances human productivity, allowing workers to perform tasks more efficiently.
3. Job restructuring – AI changes the mix of tasks within an occupation, potentially leading to new job descriptions building on different skills.

In practice, all three of these forces will likely play out differently across various sectors and economies. AI could also create entirely new job categories that we have yet to envision. As a recent National Academies study focused on the US put it, “This is a highly uncertain time for forecasting the future of work.”

Despite this uncertainty, our findings offer a grounded prediction: AI’s impact on jobs will be more muted and slower to materialize in many developing countries—especially in regions where consistent access to electricity or internet remains limited.

Policy implications: Preparing for AI’s future in developing countries

Given these insights, what steps can policymakers take to ensure that AI benefits workers in developing economies rather than bypassing them?

• Expand digital and energy infrastructure. Reliable electricity and internet access—particularly in rural areas—will be crucial for enabling AI adoption where it can be most beneficial.
• Prioritize AI augmentation over automation. Encouraging AI applications that enhance human productivity rather than replace workers can help protect livelihoods and stimulate inclusive economic growth.
• Leverage AI for healthcare and education. AI has the potential to bridge human capital shortages in critical sectors like medicine and schooling, making expertise more accessible.

While AI is transforming the global economy, its impact will not be uniform. In many developing nations, the disruption will come at a slower pace—offering a unique window of opportunity for governments and businesses to shape AI’s role in the workforce. With proactive policies, these countries can harness AI to empower workers rather than displace them.

source. Worldbank Blogs