ISLAMABAD: Genera­tive artificial intelligence (GenAI) is poised to transform labour markets worldwide, but with varying impacts across countries, according to a joint ILO and World Bank study.

The background study for the forthcoming World Development Report 2026 states that in developing economies, disruption may occur more quickly than productivity gains due to existing digital gaps and differences in how work is performed.

The study highlights a “small buffer, big bottlenecks” dynamic in developing countries. Workers in roles exposed to automation are often already connected and may face displacement pressures fairly quickly. Those in jobs with augmentation potential often lack reliable internet access, which limits their ability to achieve productivity gains.

In low- and lower-middle-income countries, the proportion of jobs vulnerable to automation is generally smaller. However, these roles often include formal, higher-quality service-sector jobs; occupations mainly held by women and younger workers; and entry-level clerical and administrative positions that traditionally served as routes into decent work.

This increases the risk of a “white-collar bypass”, where office-based jobs that supported upward mobility and women’s labour force participation in advanced economies may not fully develop in today’s developing countries, the study says.

The study investigates labour market exposure to GenAI across 135 countries, representing about two-thirds of global employment. It demonstrates that differences in digital infrastructure and job task composition are crucial in shaping how risks and opportunities are distributed between advanced and developing economies.

Exposure to GenAI is greater in advanced economies, especially in clerical and professional roles. Developing countries, although less exposed overall, face structural barriers that restrict their capacity to benefit from the technology.

The study finds that exposure to GenAI is higher in advanced economies. Around 30 to 32 per cent of employment in high-income countries is exposed, while in low-income countries, this figure is close to 1 to 15 per cent.

A key finding is the role of the digital divide. Workers in jobs vulnerable to automation are often already online, even in low-income settings, which means job losses could occur fairly quickly. These jobs tend to be relatively higher-quality in lower-income countries, including clerical and administrative roles that have historically provided a pathway to decent work, especially for women and young workers. The concern is that AI-driven automation could cut off these pathways.

Published in Dawn, April 5th, 2026