The International Labour Organization (ILO) published Gen AI: Occupational Segregation and Gender Equality in the World of Work (March 2026). The research brief draws on the ILO’s microdata covering 84 countries.
Human Level’s Take:
- Generative AI will be a big part of the future of work. It comes with advantages, like reducing workloads, increasing safety and improving worker well-being, but can also result in job loss and mental health impacts. Without ensuring adequate safeguards, it could also result in a far more unequal workforce.
- The ILO reports that female-dominated occupations like business administration and clerical work are almost twice as likely to be exposed to Gen AI compared to male-dominated fields like construction and manufacturing. They also have a higher automation risk (16% compared to 3%), putting more women at risk of losing their jobs.
- The effects vary by region, with 41% of jobs in high-income countries exposed compared to 11% in low-income countries; women face more Gen AI exposure than men in 88% of countries studied. The highest exposure of women to Gen AI is found in Pacific and Caribbean island nations, select European countries like the UK and Switzerland, and the Philippines, which have high representation of women in the services sector coupled with rapid AI growth.
- Longstanding occupational segregation, discriminatory norms, biased workplace practices, and macroeconomic policies are all at the root of this inequality. In addition, underrepresentation in STEM and AI fields can lead to embedding gender bias into AI tech, further widening the digital divide.
- Companies can help bridge the gender gap by ensuring that women workers have access to reskilling and upskilling opportunities and career counseling. At the development stage, companies can build gender equality and representation of women into the design, deployment and governance of new technologies. They can use more representative datasets to train systems and assess risks of bias to avoid embedding gender discrimination.
- These efforts need to be supported by government policies that level the playing field for women, including those that encourage women’s participation in the workforce and provide equal access to job opportunities of the future.
Some key takeaways:
- Disproportionate impacts on women workers: Generative AI (Gen AI) — AI that creates new content based on large language models — is transforming employment at an unprecedented pace, but its impacts fall unevenly along gender lines due to pre-existing inequalities in work and economic opportunity. This is important because exposure to AI can result in risks to workers, like job displacement or mental health effects. The ILO reports that female-dominated occupations like business administration and clerical work face nearly twice the exposure to Gen AI compared to male-dominated fields like construction and manufacturing (29% compared to 16%), and a far higher automation risk (16% compared to 3%). The effects also vary significantly by region and income level. Forty-one percent of jobs in high-income countries are exposed to Gen AI compared to just 11% in low-income countries. This also has implications for gender equality: women face greater Gen AI exposure than men in 88% of the countries studied, with the highest concentrations found in Pacific and Caribbean island nations, select European countries like the UK and Switzerland, and the Philippines. In large part, this is due to women's stronger representation in the services sector combined with the rapid growth of AI in these countries. Without deliberate intervention, Gen AI risks deepening existing gender inequalities.
- Root causes of inequality: What’s driving the inequality? The ILO attributes this to longstanding occupational segregation, which is sustained by discriminatory norms, biased workplace practices, and macroeconomic policies that limit women's equality of opportunity. While Gen AI is expected to create growth in tech-intensive sectors, women remain significantly underrepresented in STEM and the AI workforce (just 30% globally), which limits their ability to benefit from and shape these developments. Underrepresentation also risks embedding gender bias into AI technologies themselves, further widening the digital divide. Beyond job numbers, Gen AI's deeper impact lies in reshaping work quality: it can intensify workloads, reduce autonomy, and introduce bias, but also holds potential to ease physical demands, improve well-being, and advance workplace equality. But for women workers to benefit from these advantages, Gen AI needs to be designed inclusively and supported by strong labour market institutions and meaningful social dialogue.
- What’s needed: The ILO outlines a number of considerations to ensure that the risks of AI don’t outweigh the benefits. One key need is embedding gender equality in the design, deployment and governance of new technologies. This relies on efforts to reduce bias in development, using more representative datasets to train systems, and safeguarding against bias through bias audits or risk assessments. In addition, ensuring that women and other vulnerable groups are represented at the development stage is critical. The ILO also points to expectations for governments to address structural causes of inequality in the workplace and in unpaid care responsibilities. Labour market policies to ensure women’s access to quality jobs will need to be accompanied by care leave policies, services and infrastructure to create greater equality in the world of work.
