Artificial intelligence continues to reshape the global workforce, but its impact is far from equal across demographic lines. Recent research reveals a concerning pattern: AI automation threatens women's employment at nearly three times the rate of men's jobs, raising urgent questions about technological equity and the future of work.
The Uneven Impact of AI on Employment
A groundbreaking report from the United Nations' International Labour Organization (ILO) and Poland's National Research Institute (NASK) has uncovered a stark gender disparity in AI's workplace impact. In higher-income countries, approximately 9.6% of women's jobs face high risk of AI automation, compared to just 3.5% of positions held by men. This nearly threefold difference highlights how technological advancement may exacerbate existing workplace inequalities rather than alleviate them.
AI automation risk by gender in higher-income countries:
- Women's jobs at high risk: 9.6%
- Men's jobs at high risk: 3.5%
Administrative Roles in the Crosshairs
The disproportionate risk to women's employment stems largely from occupational segregation patterns. Clerical and administrative positions—traditionally dominated by women—show the highest exposure to AI automation capabilities. Between 2000 and 2019, women held between 93% and 97% of secretary and administrative assistant roles in the United States, while comprising only 40-44% of the overall workforce. These administrative positions rank as the fifth most common profession for American women according to Department of Labor statistics.
Administrative positions in the US (2000-2019):
- Percentage held by women: 93-97%
- Women in overall workforce: 40-44%
Automation versus Augmentation
Harvard Business School associate professor Rembrand Koning frames the challenge through two distinct lenses: automation versus augmentation. We can think of this as a threat, which is that it's going to automate away a lot of these clerical jobs that might be held more by women, Koning explained. On the other hand, we can think of AI as automating a lot of this work, allowing workers to take on tasks that might be higher paying. This perspective suggests AI could either eliminate jobs or transform them into more valuable roles—with the outcome depending largely on implementation approaches.
Gender Differences in AI Adoption
Compounding the automation risk is a troubling adoption gap. Koning's research reveals women use AI tools at rates approximately 25% lower than their male counterparts. This disparity appears linked to ethical concerns and workplace dynamics—women often worry about being perceived as cheating by using AI or having their intelligence questioned by male colleagues. Meanwhile, men tend to display greater confidence—perhaps overconfidence—that AI use will benefit their careers without negative repercussions.
AI adoption gap:
- Women use AI tools at approximately 25% lower rates than men
The Leadership Responsibility
Addressing these gender disparities requires proactive leadership intervention. Rather than placing the burden on individual women to overcome adoption barriers, Koning emphasizes that workplace leaders must establish clear expectations and resources around AI use. This becomes particularly crucial in environments where AI experimentation happens informally and is dominated by male employees. If we want to make sure it's inclusive, it includes all workers, it's the job of a leader to bring everybody in, Koning noted.
The Broader Context of AI Safety
Beyond gender impacts, AI safety emerges as a critical societal concern. As highlighted in a recent Imagination in Action panel discussion, AI safety isn't merely about preventing science-fiction scenarios of machine takeover. Instead, immediate risks involve algorithmic bias affecting access to housing, jobs, credit, and even legal outcomes. Panelist Albert Cahn pointed to real-world examples like the Midas insurance fraud detection system that incorrectly flagged innocent individuals, creating significant hardship.
Continuous Measurement and Governance
Experts emphasize that ensuring AI safety requires ongoing vigilance. As panelist Cam Kerry noted, In carpentry, the maxim is 'measure twice, cut once.' When it comes to AI, it has to be 'measure, measure, measure and measure again.' This continuous assessment process must extend from system development through deployment and beyond. Organizations like the National Institute of Standards and Technology are developing measurement frameworks, but these efforts need significant scaling to match AI's rapid advancement.
Moving Forward Responsibly
As AI continues transforming workplaces globally, addressing its uneven impact becomes increasingly urgent. The technology's potential to either exacerbate or reduce inequalities depends largely on how it's implemented and governed. By recognizing AI's disproportionate effects on women's employment, establishing inclusive adoption practices, and implementing robust safety measures, organizations can help ensure technological progress benefits all workers equitably rather than reinforcing existing disparities.