AI threatens clerical jobs in developing countries, widening inequality
Generative AI will displace workers in clerical and administrative roles across low-income countries faster than it creates productivity gains, according to a joint analysis by the International Labour Organization and World Bank. These positions represent some of the few quality jobs available in poorer economies and have historically served as entry points for women and young people into stable work.
The risk stems from a stark digital divide. Workers in jobs vulnerable to automation already have internet access and face displacement pressure quickly. Workers positioned to benefit from AI tools often lack reliable connectivity, preventing them from gaining productivity advantages.
Exposure varies sharply by country wealth
Advanced economies show the highest AI exposure: 30% or more of jobs in the US and France face potential AI impact. Lower-income countries like Ethiopia and Zimbabwe see exposure rates below 10%.
Most exposure globally-about 17% of jobs-falls into the "augmentation" category, where AI supports rather than replaces work. Only 8% of jobs face higher automation risk. This suggests developing countries could theoretically gain more from productivity gains than they lose to job displacement.
The reality is different. Of 441.8 million jobs with augmentation potential across countries studied, around 66.9 million workers lack internet access. That's unrealised productivity potential concentrated in poorer regions.
The "white-collar bypass" problem
In low- and lower-middle-income countries, automation exposure concentrates in formal service-sector roles-exactly the jobs that have historically offered pathways to decent work. Women and younger workers are overrepresented in these occupations.
Office-based clerical and administrative positions supported upward mobility in advanced economies over decades. That same job creation pathway may not materialise in developing countries if AI automates these roles before they fully develop.
Job titles mask real differences in task content
Standard measures of AI exposure overestimate risk in developing countries. The same job title can involve different work depending on location.
Workers in developing economies perform fewer non-routine analytical tasks within the same occupations classified as highly exposed elsewhere. When researchers adjusted exposure scores to reflect actual task content rather than job titles alone, many developing countries moved lower in automation risk rankings.
This matters for HR planning. A "financial analyst" role in a low-income country may involve more routine data entry than the same title in an advanced economy-changing both the automation risk and the skills needed to adapt.
Who faces pressure first
Exposure increases with education and household income across all countries. Women and younger workers face higher automation exposure in upper-middle and high-income countries.
In poorer countries, the digitally connected segment of skilled workers-often a small group-may experience labour-market pressure earlier than broader populations.
What HR leaders should prepare for
GenAI's impact depends on more than technology alone. Digital infrastructure, how work is organised within roles, education systems, and labour protections all shape outcomes.
For HR teams in organisations operating across income levels, this means:
- Expanding digital connectivity is essential to enable productivity gains, not just automation
- Skills upgrading programs should focus on non-routine analytical and digital tasks where workers lack experience
- Labour protections and social safety nets become critical for managing displacement in vulnerable roles
- Gender-aware policies are needed for clerical and administrative occupations where women concentrate
Without investment in digital infrastructure and active workforce policies, productivity gains will flow primarily to richer countries and better-equipped organisations. For connected workers in poorer regions, displacement risk arrives before opportunity.
HR leaders managing global workforces should consider how these dynamics affect recruitment, retention, and skills development across different markets. AI for Human Resources training can help teams understand these labour-market shifts and plan accordingly. For HR executives, an AI Learning Path for CHROs addresses workforce analytics and talent strategy in the context of automation decisions.
Your membership also unlocks:
