AI Will Hit Rich and Poor Countries Very Differently — ILO-World Bank Study of 135 Nations

Two-thirds of the world's workers just got measured for their AI exposure — and the results split neatly along the global wealth line. [Fact] A new joint study by the International Labour Organization and the World Bank, covering 135 countries and roughly two-thirds of global employment, finds that generative AI's impact on jobs varies dramatically depending on where you live and what you do.

That's not a vague prediction. It's the empirical backbone of the upcoming World Development Report 2026, and the numbers tell an uncomfortable story about who stands to gain, who stands to lose, and who might get left out entirely.

Rich Countries: The Clerical and Professional Squeeze

If you work a desk job in a high-income country, this paper is essentially about you. [Fact] Advanced economies show the highest GenAI exposure rates, concentrated in clerical, administrative, and professional occupations — exactly the roles where AI language tools are already making inroads.

Think about what that means in practice. Administrative assistants scheduling meetings, data entry workers processing forms, accountants reconciling ledgers — these are tasks that large language models can already handle or will handle soon. The exposure isn't theoretical. Companies are actively piloting these tools right now.

But here's the nuance that often gets lost in the headlines. [Claim] The ILO-World Bank researchers argue that high exposure doesn't automatically mean high replacement. In wealthy nations, workers tend to have the education, digital literacy, and institutional support to adapt. The risk is real, but so is the capacity to respond.

Developing Countries: Disruption Before Benefit

This is where the study gets genuinely worrying. [Fact] In developing economies, the paper finds that AI-driven disruption may arrive before the productivity benefits do. That's a brutal sequence.

Here's why. Workers in lower-income countries tend to perform fewer non-routine analytical tasks — the kind where AI augmentation actually boosts productivity. Instead, their jobs skew toward routine and manual tasks. So when AI does reach these economies, it's more likely to displace than to enhance.

And it gets worse. [Fact] The study highlights a critical digital divide: many workers in developing countries lack basic internet access, which means they can't even use AI tools that might help them adapt. You can't benefit from a technology you can't reach.

[Claim] The researchers warn that without deliberate policy intervention, generative AI could widen the gap between rich and poor nations rather than closing it. The countries with the least capacity to absorb economic shocks are the ones most likely to face them.

Women and Young Workers Bear Disproportionate Risk

[Fact] The study specifically flags women and youth as disproportionately vulnerable to AI disruption across all country income levels. Women are overrepresented in clerical and administrative roles — the very occupations with the highest AI exposure. Young workers, meanwhile, are more likely to hold entry-level positions where routine tasks dominate.

This isn't just an economic concern. When disruption hits the workers who already face structural disadvantages, it compounds existing inequality rather than just creating new kinds. A young woman working as a customer service representative in a developing country faces triple exposure: her occupation, her demographic, and her country's limited digital infrastructure.

What the 135-Country Data Actually Shows

Let me put the scale in perspective. [Fact] This isn't a model built on assumptions about a handful of economies. The ILO-World Bank study spans 135 countries, representing roughly two-thirds of global employment. That makes it one of the most comprehensive assessments of AI's labor market impact ever published.

[Fact] The findings feed directly into the World Development Report 2026, the World Bank's flagship annual publication. When these two institutions collaborate at this scale, policymakers tend to pay attention — and that's exactly the point. The paper is explicitly designed to inform national AI strategies and labor market policies.

The core message is clear: a one-size-fits-all approach to AI and employment policy will fail. What works for software developers in Stockholm won't work for garment workers in Dhaka. The policy response needs to be as differentiated as the impact itself.

What This Means for You

If you're reading this from a high-income country, the practical takeaway is straightforward: the clerical and administrative squeeze is accelerating. Skills that complement AI — judgment, interpersonal nuance, creative problem-solving — are your best hedge. If your job is mostly about processing information in predictable patterns, the timeline for disruption just got more concrete.

If you're in a developing economy, the picture is harder. Access to digital tools, internet connectivity, and retraining programs matters enormously. [Claim] The ILO-World Bank authors argue that international cooperation on digital infrastructure and education investment is not optional — it's essential to prevent AI from becoming another engine of global inequality.

And regardless of where you are, if you're a woman or early in your career, pay extra attention to diversifying your skills beyond routine tasks. The data isn't subtle about who faces the most pressure.

Sources

1. ILO / World Bank, "New ILO-World Bank paper highlights uneven global impact of generative AI on jobs" (March 2026). Link

Update History

• 2026-04-15: Initial publication based on ILO-World Bank joint study for WDR 2026

---

This analysis was produced with AI assistance (Claude). All statistics are sourced from the ILO-World Bank joint study on generative AI and labor markets across 135 countries. For occupation-level data, see individual occupation pages.