Generative AI May Disrupt Jobs in Developing Nations Before Delivering Gains

A global study by the World Bank, ILO, and NASK finds that generative AI will affect more jobs in richer countries, but developing nations may face disruption first due to weaker digital infrastructure. The research warns that without better internet access and skills development, AI could deepen global inequality instead of boosting productivity everywhere.

Generative artificial intelligence is rapidly transforming how work is done worldwide. But its impact will not be the same everywhere. A new study by the World Bank’s Poverty and Equity Global Department, conducted with researchers from the International Labour Organization (ILO) and Poland’s National Research Institute NASK, shows that while AI could boost productivity globally, developing countries may experience its disruptions before they see the benefits.

The research examines how generative AI could affect jobs in 135 countries. By analyzing employment patterns, workplace tasks, and internet access, the study finds that the effects of AI will depend heavily on how jobs are structured and whether workers have the digital tools needed to use the technology.

Why Richer Economies Are More Exposed

One of the key findings is that AI exposure increases as economies develop. In high-income countries, around one-third of jobs could be affected by generative AI. In low-income countries, the share is closer to fifteen percent.

The difference mainly comes from the types of jobs people do. Wealthier economies rely more on office-based and knowledge-intensive work such as administration, management, finance, and professional services. These jobs involve writing, analyzing data, and processing information, tasks that generative AI systems can assist with or perform.

In contrast, many workers in poorer countries are employed in agriculture, construction, or other manual sectors where AI tools have limited direct application. This means fewer jobs are immediately exposed to the technology.

Opportunity or Automation?

Not all AI exposure means job loss. In many cases, generative AI is expected to support workers rather than replace them. Tools such as AI writing assistants or data analysis systems can help employees complete tasks faster or improve decision-making.

These “augmentation” effects could increase productivity in many industries. For example, workers in customer service, business services, or management roles may use AI to draft reports, summarize information, or assist with communication.

However, some occupations face greater automation risk. Clerical jobs, administrative roles, and certain professional services involve routine digital tasks that AI systems can increasingly perform on their own. This means some office jobs could be reshaped or reduced as AI adoption grows.

The Digital Divide Slows the Benefits

The study highlights a major challenge: the global digital divide. Generative AI tools require stable internet connections and access to digital devices. In many developing countries, these basic requirements are still not widely available.

As a result, workers who could benefit from AI assistance often cannot access the technology. At the same time, workers whose jobs could be automated may still face disruption because companies can adopt AI systems remotely or through global platforms.

This imbalance creates a situation where disruption may arrive before productivity gains. In wealthier countries, most workers in AI-exposed jobs already have reliable internet access. In poorer countries, many workers remain offline, preventing them from using the same tools.

Who in the Workforce Is Most Affected

The research also shows that exposure to generative AI varies across workers and industries. Clerical and administrative roles face some of the highest risks because many of their tasks involve writing, documentation, and information processing.

Women may be particularly affected because they are more likely to work in these types of jobs. Younger workers may also face higher exposure since entry-level office positions often involve routine tasks that AI can assist with or automate.

Exposure also increases with education levels. Workers with university degrees are more likely to perform analytical and digital tasks that AI systems can support.

Certain industries are more affected than others. Financial services, business services, and communication sectors show high levels of exposure because they depend heavily on digital work. Agriculture remains largely unaffected because most tasks require physical activity and local knowledge.

Why Job Tasks Matter

Another important finding is that the same job can involve different tasks in different countries. For example, a clerical job in a high-income country may involve advanced computer use and analytical tasks. In a lower-income country, the same job may include more routine procedures or manual components.

Because generative AI works best with digital and analytical tasks, these differences matter. When researchers adjusted their estimates to reflect actual job tasks, many developing countries appeared less exposed than standard job classifications suggested.

This means job titles alone do not determine how AI will affect workers. What people actually do in their jobs, and the technology available to them, plays a major role.

A Technology That Could Widen Inequality

Overall, the study concludes that generative AI could deepen existing global inequalities if countries do not invest in digital infrastructure and workforce skills. Productivity gains are likely to appear first in economies with strong internet access, digital systems, and highly skilled workers.

For developing countries, the risk is that AI could reshape certain office jobs before those roles fully expand, limiting opportunities for young workers and women who often rely on these positions as entry points to stable employment.

The researchers emphasize that the outcome is not predetermined. With the right investments in connectivity, education, and digital skills, generative AI could help workers improve productivity and create new opportunities.

But without those efforts, the technology may reinforce the gap between countries that can use AI effectively and those that cannot.