Women’s jobs face twice the automation risk as men’s: ILO

Generative artificial intelligence is poised to reshape global labour markets, but new data from the International Labour Organisation confirms that women will bear a disproportionate share of the disruption, with female-dominated occupations facing nearly twice the exposure risk of male-dominated roles. The findings, published in the research brief titled “GenAI, Occupational Segregation and Gender Equality in the World of Work”, reveal that approximately 29 per cent of female-dominated occupations are exposed to GenAI automation, compared with just 16 per cent of male-dominated jobs.

The disparity becomes even more pronounced at the highest levels of automation risk. While 16 per cent of female-dominated occupations fall into the top exposure categories, only three per cent of male-dominated roles face equivalent risk. This gap is not accidental but stems from deep structural patterns of occupational segregation that have persisted for decades. Women remain heavily concentrated in clerical, administrative and business support positions, including roles such as secretaries, receptionists, payroll clerks and accounting assistants. These positions involve routine, codifiable tasks that AI systems can replicate with relative ease. Men, by contrast, dominate sectors such as construction, manufacturing and manual trades, where physical labour and non-routine tasks offer greater insulation from automation.

The ILO analysis examined country-level data and found that women face higher GenAI exposure than men in 88 per cent of countries studied. In several economies, more than 40 per cent of women’s employment is exposed to the technology. High-income countries including Switzerland, the United Kingdom and the Philippines, along with small island developing states in the Caribbean and the Pacific, show particularly elevated exposure rates. Overall, 41 per cent of jobs in high-income countries are exposed to GenAI, compared with just 11 per cent in low-income countries, reflecting differences in occupational structures and digital readiness.

Anam Butt, co-author of the ILO research, stated that “generative AI is not entering a neutral labour market.” She explained that social norms, unequal care responsibilities and labour policies continue to shape who enters which occupations, leaving women concentrated in roles more vulnerable to automation. This concentration is compounded by women’s underrepresentation in the fields that are shaping AI itself. Women constitute only around 30 per cent of the global AI workforce, a figure that has barely changed since 2016. When women are absent from AI development and decision-making processes, the technology tends to reproduce existing societal biases.

The report emphasises that GenAI systems are not inherently neutral. AI technologies can replicate and amplify gender biases, disadvantaging women in recruitment, pay decisions, credit scoring and access to services. The impact is further compounded for women who face intersecting forms of discrimination based on race, disability or migration status. The ILO stresses that while much public discussion focuses on job displacement, the more immediate impact of generative AI is likely to fall on job quality, including changes to tasks, intensified workloads and increased workplace surveillance.

Historical context helps illuminate why women find themselves in this vulnerable position. Occupational segregation by gender has roots stretching back to the Industrial Revolution, when women were channelled into textile and garment work, later moving into clerical positions as the service economy expanded. The 20th century saw women enter office work in large numbers, particularly during the World Wars when they filled administrative roles vacated by men. However, these positions were often classified as secondary and paid less, a pattern that persisted even as women’s labour force participation increased dramatically in the latter half of the century. By the time digital technologies began transforming workplaces in the 1980s and 1990s, women were already concentrated in the routine information-processing roles that would prove most susceptible to computerisation and, now, AI automation.

The current wave of generative AI tools, including ChatGPT and similar systems, represents a significant escalation in automation capabilities. Unlike earlier technologies that automated physical tasks or simple calculations, GenAI can generate text, analyse documents, manage communications and perform complex administrative functions. These are precisely the tasks that dominate female-concentrated occupations. The technology is transforming how work is carried out across industries, particularly in roles involving repetitive and administrative duties.

Policy responses will determine whether GenAI becomes a tool for equality or entrenches existing disparities. Janine Berg, Senior Economist and co-author of the report, stated that “women are at higher risk, but the outcome is not predetermined.” She added that “with the right policies, social dialogue and gender-responsive design, we can prevent AI from reinforcing existing discrimination.” The ILO calls for embedding gender considerations in AI design, improving women’s access to STEM education, and strengthening labour market institutions as critical steps to ensure the technology benefits all workers.

The organisation urges governments, employers and workers to collaborate on how GenAI is introduced in workplaces, ensuring that technological change supports productivity, job quality and a more inclusive future. Without deliberate intervention, the rapid adoption of artificial intelligence risks widening gender inequalities that have persisted in labour markets for generations. The data suggests that the window for proactive policy is narrow, as adoption of AI tools accelerates across high-income economies and begins to penetrate developing markets.