A Tale of Two Automations
For years, the narrative around AI and employment has been one of universal disruption. Yet, the World Bank’s latest analysis argues for a more nuanced view, suggesting that the threat is not uniform. The report finds that jobs in high-income countries
are more than three times as likely to be at risk from generative AI automation than those in low- and middle-income countries. Specifically, 14.2% of jobs are considered at risk in developed economies, compared to just 4.5% in developing ones. This is because the types of work prevalent in these economies are fundamentally different. The conversation in developed nations centres on replacing cognitive tasks, while in developing countries, where manual and agrarian labour is more common, the immediate threat is lower. The report's core thesis is that we are not facing one automation wave, but two distinct challenges that require separate solutions.
High-Skill vs. High-Touch Jobs
In wealthy countries, the jobs most exposed to AI are cognitive, non-routine roles—the very ones that have historically been considered secure, white-collar professions. These include tasks related to analysis, administration, and even some creative fields. The risk here is the hollowing out of the professional middle class, as AI begins to perform tasks previously done by skilled workers. Paradoxically, this also creates an opportunity to augment productivity, with 18.7% of jobs in these economies poised to see a significant boost from AI integration. In contrast, developing economies are dominated by sectors like agriculture and small enterprises, where jobs are more manual and less susceptible to immediate automation by current AI models. For these nations, AI’s biggest promise isn't replacing workers, but amplifying their capabilities—what the report calls augmentation. For example, AI can provide scarce expertise to doctors, farmers, and teachers, radically speeding up development.
The Ladder to Development at Risk
While the immediate job displacement risk is lower in developing nations, the report issues a stark long-term warning. Historically, countries have moved from agriculture to manufacturing and then to services, with each step creating more productive jobs. AI threatens this traditional development ladder. Key entry-level service jobs that have been a gateway to the middle class—such as those in call centres and back-office processing—are highly vulnerable to automation. The report cautions that this could close off a crucial route to economic mobility for millions, potentially widening the gap between countries. If developing economies fail to adapt, they risk missing the AI revolution entirely, much like many missed the first Industrial Revolution and spent centuries trying to catch up.
The Infrastructure and Skills Divide
The capacity to respond to these divergent risks also differs enormously. High-income nations can focus on retraining and social safety nets for displaced professionals. However, many developing countries lack the fundamental infrastructure to even participate in the AI economy, let alone manage its downsides. According to the World Bank, as of 2024, 2.6 billion people remain offline, and internet usage in low-income countries is just 27%, compared to over 90% in high-income ones. Without reliable electricity, affordable internet, and widespread digital skills, harnessing AI's benefits is impossible. The report stresses that before developing nations can think about frontier AI models, they must invest in these foundational pillars and focus on adapting low-cost, practical AI tools to local languages and contexts.
