World Bank: In rich countries, jobs are more exposed to artificial intelligence

The factor of labor market composition

A second point raised by the institution concerns low-income countries instead. In these contexts, where agriculture and informal work account for a larger share of total employment, direct exposure to AI appears more limited. The World Bank warns, however, that this does not amount to immunity: the impact could materialize through export markets, investment flows, and the spread of tools accessible via the cloud even without large local infrastructure.

The World Bank's chief economist, Indermit Gill, reiterated that the opportunities offered by AI are not limited to advanced economies. "Sie benötigen keine großen Modelle oder Big-Data-Zentren, um von KI zu profitieren", he said, with reference to lower-income countries. According to Gill, access to lightweight models, ready-to-use applications, and local-language services can allow even African or South Asian businesses to integrate artificial intelligence into their production processes.

Opportunities for low-income countries

The report also draws attention to gaps in preparedness. In rich countries, where the workforce is on average more educated and digitalized, professional reskilling can count on already structured training systems. In poorer economies, however, the absence of continuous training programs risks leaving workers without tools to adapt, even if AI does not directly automate their tasks. The World Bank therefore urges governments to invest in active labor policies and basic digital infrastructure.

Among the countries cited as most exposed are the United States, where the professional and financial services sector is historically developed, and South Korea, with an advanced manufacturing industry and a strong presence of large technology groups. The report also mentions intermediate cases such as Mexico, where exposure varies significantly between formal and informal sectors. For lower-middle-income countries, the institution flags the need to monitor the impact on manufacturing sectors linked to global value chains.

Recommendations to governments

The document also reviews situations of specific vulnerability, such as Egypt, where large segments of the active population are employed in public services and tourism, and Ethiopia, where the rural occupational structure limits direct exposure but not indirect exposure. The reference to Afghanistan, on the other hand, is used by analysts to illustrate an extreme case in which the absence of updated labor market data makes any assessment of the impact of AI on employment more difficult.

The World Bank places its analysis within the framework of preparatory work for the next development summit, in which the themes of technological innovation and the digital transition are expected to take a central role. The stated goal is to prevent the spread of AI from amplifying existing inequalities between rich and poor countries, and to ensure that the benefits of the new technology are not concentrated in a few advanced economies.

On the operational side, the institution recommends that low-income countries focus investments on connectivity, reliable electricity, and basic digital skills, rather than on building large data centers. For high-income countries, the indicated priority is rather regulation, algorithmic transparency, and the strengthening of social protection systems, so as to manage the transition without large-scale occupational shocks.

According to World Bank analysts, the picture that emerges is twofold: on the one hand, rich countries must prepare for a potential restructuring of the labor market; on the other, poorer countries have the opportunity to adopt artificial intelligence solutions at low cost, provided they are accompanied by coherent public policies. Gill concluded by recalling that the speed of AI adoption will depend largely on the regulatory and training choices of individual governments, rather than on the availability of computational resources.