May 21, 2025
By Rehan Piracha
LAHORE: A major new report from the International Labour Organization (ILO) reveals that generative artificial intelligence (GenAI) is poised to disrupt labor markets worldwide—with women facing significantly higher risks of job automation than men.
According to ILO Working Paper 140: Generative AI and Jobs: A Refined Global Index of Occupational Exposure (2025), 9.6% of female employment in high-income countries (HICs) is in occupations at the highest risk of automation, compared to just 3.5% of male employment—a disparity nearly three times as large.
The study, which refines the ILO’s 2023 analysis with expanded datasets and expert evaluations, provides the most comprehensive assessment yet of how AI could transform jobs across industries, regions, and demographic groups. Its findings highlight not only the uneven impact of automation but also the urgent need for policies that protect vulnerable workers—particularly women in clerical and administrative roles—from being left behind in the AI-driven economy.
Women bear the brunt of automation risks
The study categorizes occupations into four “exposure gradients” based on how susceptible their tasks are to automation by GenAI:
Gradient 4 (Highest Exposure): Jobs where most tasks can be fully automated (e.g., data entry clerks, typists, accounting clerks).
Gradient 3 (High Exposure): Jobs where many tasks can be automated, but human oversight remains necessary (e.g., financial analysts, translators).
Gradients 1–2 (Low to Moderate Exposure): Jobs where only some tasks are automatable (e.g., managers, sales workers).
Globally, 24% of all employment falls into one of these exposure categories, but the risks are not evenly distributed. Women are overrepresented in high-exposure jobs, particularly in clerical and service roles. In HICs, 41% of female workers are in occupations with some level of GenAI exposure, compared to 28% of male workers.
The gender gap widens at the highest risk level: While 4.7% of women worldwide are in Gradient 4 jobs, only 2.4% of men are. In HICs, this gap expands to 9.6% of women vs. 3.5% of men.
“Occupational segregation is driving this divide,” explains Pawel Gmyrek, the study’s lead author. “Women are disproportionately concentrated in administrative, customer service, and data-processing roles—exactly the kinds of jobs GenAI can most easily automate.”
High-income countries face highest exposure
The study finds that 34% of employment in HICs is exposed to GenAI, compared to just 11% in low-income countries (LICs). This gap reflects differences in economic structure, digital infrastructure, and job composition. In HICs, digitized office jobs (e.g., secretarial work, financial analysis) dominate, making them more susceptible to AI-driven automation.
In LICs, agriculture and manual labor remain prevalent, shielding workers from immediate AI disruption—though this could change as technology spreads.
However, even within HICs, exposure varies. Europe and Central Asia have the highest female exposure (39%), while the Americas show a narrower gap (30% of women vs. 20% of men exposed).
Digitised jobs are more at risk
The 2025 study revises earlier estimates based on real-world AI advancements. Some clerical jobs (e.g., general office clerks, typists) saw lower automation scores than in 2023, as experience showed GenAI struggles with nuanced tasks like handling sensitive documents or complex customer interactions.
However, highly digitized roles (e.g., web developers, statisticians, financial advisers) saw increased exposure, as AI’s ability to generate code, analyze data, and produce reports improved dramatically. “AI is evolving faster than we anticipated,” says Janine Berg, ILO Senior Economist and co-author. “Tasks we thought were safe in 2023 are now automatable, while others remain stubbornly human-dependent.”
Why are women more exposed?
The gender disparity in AI exposure is not accidental—it reflects long-standing inequalities in workforce participation.
Women are overrepresented in routine cognitive jobs (e.g., data entry, customer support) and underrepresented in manual or creative roles (e.g., construction, engineering). Since GenAI excels at text and data processing, female-dominated jobs are at greater risk.
Many women work in part-time, temporary, or informal jobs with fewer protections. Employers may see these roles as low-cost candidates for automation, leaving women with fewer alternatives.
While women in HICs often have strong digital literacy, those in developing economies face limited access to tech training. Without upskilling, they risk being excluded from new AI-augmented roles.
“This isn’t just about job losses—it’s about whether women will have access to the better jobs AI creates,” warns Laura Nurski, an AI labor market expert at the Centre for European Policy Studies.
The study emphasizes that most jobs will transform rather than disappear, as few occupations are fully automatable. However, the transition could still be painful without intervention.
If displaced workers flood into low-paying sectors, wages could stagnate—particularly for women re-entering the workforce after automation. Countries with weak education and social protection systems may see rising inequality, as workers without AI skills fall behind.
How to ensure a fair AI transition?
The ILO urges governments, employers, and unions to take proactive steps with targeted reskilling for women. This includes digital literacy programs tailored to female workers in high-risk jobs as well as incentives for women to enter STEM and AI-related fields, where demand is growing.
The ILO called for strengthening labour protections with expansion in social safety nets for displaced workers. In addition, it asked employers to regulate AI use in hiring to prevent gender bias in recruitment and redesign jobs to combine AI efficiency with human judgment (e.g., AI-assisted paralegals, not fully automated ones).
The ILO study concludes that AI’s impact depends on policy choices. Without intervention, automation could deepen gender and income inequality. But with the right strategies, AI could also create new opportunities and reduce drudgery in female-dominated jobs.
“The key question isn’t whether AI will change work—it’s who will benefit from those changes,” says Gmyrek. “If we act now, we can ensure that women aren’t left behind in the next industrial revolution.”
The ILO will expand its research with country-level studies, starting with Poland, to refine policy recommendations. Meanwhile, governments and businesses must prioritize equity in the AI transition—before the gender gap grows even wider.