Is AI excluding women from top corporate roles?
The newest threat to the ‘leaky pipeline’ is AI, adding a new systemic barrier to equality at the workplace
Last week, global consultancy firm Capgemini published a report on the state of diversity, equity and inclusion (DEI) in the corporate sector, based on a survey of 2,750 leaders across 11 countries. The document, titled Gender and Leadership: Navigating Bias, Opportunity and Change, revealed several positive highlights.
A majority of the respondents (77%) believe that women are as effective as men in leadership roles. Over 65% said that having women in leadership roles improves business outcome, and 58% women and 59% men cited confidence as their key strength behind their success.
Since the survey featured only three people who identify as non-binary, the results don’t affirm an improvement of DEI metrics across the board. But, as far as women’s empowerment in the workplace is concerned, it seems attitudes are shifting—even though the gender gap persists in key parameters such as pay parity (52% of the women said they were paid unfairly, while 40% men believed they received a distinct pay advantage).
The report, however, made headlines for calling out a bias that isn’t as visible or obvious. Half of the male respondents considered future-ready critical leadership skills—such as the use of artificial intelligence (AI), automation, data analysis, innovation—to be “inherently masculine". In contrast, two-thirds of the women saw these skills as gender neutral, though AI proficiency ranked low across genders (47% men and 45% women mentioning AI proficiency to be one of their key skills).
In a world where women make up only 28% of the STEM workforce (as per a World Economic Forum survey from 2024), such biases are increasingly likely to cut women out of leadership roles.
With AI tilting the scales in favour of men, it’s worth asking what AI researcher Chen Yu did in a paper last year: “Will AI serve as a great equalizer, or will it become a tool that entrenches the gender divide deeper into the societal bedrock?"
To begin with, the underrepresentation of women in STEM is one of the main reasons these biases are inherently built into AI systems, which are trained on dominant modes of communication, thinking and behaviour.
In 2018, Amazon’s AI-based recruitment tool was exposed as having gone rogue, especially after it kept rejecting women applicants. It turned out that the algorithm had been trained with data for the preceding 10 years and led to believe that men are more suitable for IT jobs over women. Every time the machine detected words like “women’s college" or “women leader" on a résumé, it would unceremoniously dump it.
Since then, improvements have been made to machine learning to reverse such biases and reduce inequality through gender-neutral hiring algorithms. Gender bias audits are a mandate for major businesses like Deloitte, Google, and Microsoft these days. But a so-called gender audit will only remain a numbers game unless the stakeholders bring intersectionality into the bigger picture.
Findings from McKinsey’s Women in the Workplace Report 2024 indicate that parity for white women in corporate roles comes faster than it for women of colour. If it takes 22 years for the former to reach parity, the latter needs 48 years to get to the same level.
Yu’s research also pointed out regulatory oversight on intersectional impacts—how race, class, and other identities intersect with gender in shaping technological outcomes.
Women globally occupy only about 22% of AI-related roles. Earlier this year, researchers Kanta Singh and Antara Lahiri wrote in an article for The Hindu that women’s representation in the STEM workforce in India remains disproportionately low at 27%, despite women constituting 43% of STEM graduates in the country. The infamous “leaky pipeline" of women talent in the workforce is widely attributed to systemic barriers, biases, sexism and cultural factors (such as lack of support for family and work-life balance). The newest addition to this list now is AI readiness.
Even if we set aside leadership roles for a moment, the threat of AI-driven job displacement disproportionately impacts women at all levels in the workforce. According to a UN report, nearly 28% of jobs typically held by women are vulnerable to AI and automation, compared to 21% for men, mainly because women dominate sectors like healthcare, education, and clerical work, areas that are highly susceptible to automation. A recent International Labour Organization (ILO) study highlights that women in administrative and clerical roles are three times more likely than men to lose jobs due to AI.
Even as businesses embrace AI’s potential to transform outcomes, without gender inclusivity, this technological disruption risks deepening inequality in the workforce. Addressing these disparities will require systemic change—spanning education, policy, technology design, and social norms—to ensure that AI becomes a driver of equitable growth rather than a catalyst of exclusion. A technological solution to bias by training algorithms is just the tip of the iceberg.
Work Vibes is a fortnightly column on ideas to help you thrive at what you do.
topics
