'A relatively higher number of women will need to transition from disrupted roles and relatively fewer currently have the skills to do so'
While artificial intelligence (AI) is being embraced to tackle a variety of challenges, it's also deepening the gender gap in the workforce, a report from the World Economic Forum finds.
The report, produced in collaboration with LinkedIn, found that 57% of women are more likely to be in roles disrupted by AI, compared to 43% of men.
Further, fewer women than men will see their work augmented (46% vs 54%) by GenAI.
“A relatively higher number of women will need to transition from disrupted roles and relatively fewer currently have the skills to do so,” the report states.
Without proper implementation and diverse deployment, the report warns that advancing AI with limited talent diversity could lead to "economic drag and AI-driven inequality."
To ensure women benefit from AI-driven automation and don't fall further behind —the report calls for the workforce to prioritize efforts in upskilling, fair hiring, and equitable performance evaluation.
Industry leaders are urged to actively integrate women into AI leadership and development, and policymakers are urged to create frameworks that support inclusive upskilling and fair workplace integration.
Without progress on gender parity, the report says, organizations risk limiting their “capacity for innovation and long-term competitiveness”.
For Annie Theriault, a managing partner at Cross Border Impact Venture—a women-led health technology investment firm—part of bridging the gender gap means empowering employees to embrace AI. This is particularly true for women who may be mid-career professionals and new to the technology.
She explains that employers should “demystify conversations with employees—not just about how AI is expected to be used in jobs, but also about the potential job losses that may come with this technological shift.
“Leadership at these organizations is modernizing using AI for their shareholders, and then they quietly say it within the organization,” she says. “Nobody is served by industries being coy or quiet about the fact that job losses are expected in the next decade or two.”
A report published in 2024 found that, on average, only 15 per cent of workers aged 45 to 65 across Europe and the U.S. reported using AI in the workplace. Most of those who did adopt AI tools were self-taught. Many workers remain hesitant to adopt AI, the study finds.
A 2024 PwC report on global gender trends found that women had a less clear understanding of how AI would affect roles, making them less prepared for an AI-driven economy.
But AI isn't just creating this gender gap—it seems to be amplifying a long-standing inequality in leadership and STEM fields.
The World Economic Forum report found while women held 24.2 per cent of managerial roles in STEM, only 12.2 per cent made it to C-suite roles.
“It is a lack of recognition of talent and an inability to appropriately facilitate that talent,” says Lisa Willis, assistant professor of biological sciences at the University of Alberta.
Willis notes that one reason for the lack of representation in the AI field—and in STEM overall—is due to how hostile and unsupportive the environment can be.
“Women come in with new ideas. They want to do things differently, and they have different goals,” she says. “They don’t feel like their thoughts and opinions are valued.”
She points to a structural problem where women enter the field with new ideas but aren’t granted the autonomy or support to pursue them.
The report notes that past research has shown employers prioritize upskilling men over women in most countries, with exceptions like Belgium and India. It’s a major barrier that stops women in these fields from being more ambitious, she says.
Although women's presence in tech fields has grown over the years, the report finds that women still represent less than one-third of the STEM workforce.
Willis and her colleagues have identified 12 principles that organizations can adopt to reduce bias and discrimination against marginalized communities in STEM.
One of those principles includes “effective” mentorship, which Willis finds crucial in helping women in AI and STEM reach leadership roles.
“When you are part of the out-group, you don't have access to the knowledge about how to advance your career. And so, opportunities come up, and you don't hear about them in time,” she says.
Having a mentor in a leadership role—someone who actively goes out of their way to support others, share opportunities, and provide transparency around skills and career development—can help women move up faster , Willis explains.
But how can organizations also make sure they're making progress in bridging these gaps? Willis says it's through examining both data and qualitative insights.
She shares the example of a gas and oil company that developed a dashboard for every level of the organization, tracking the percentage of male and female employees, as well as how long each person had been in their role. This allowed them to monitor progress against internal diversity targets.
“They've also created training for managers on how to recognize excellence and develop a more holistic view of what excellence means,” Willis says. The training also included guidance on identifying unconscious bias and discrimination.
Beyond data, she recommends gathering feedback through interviews and staff surveys.
“Numbers are important because they help tell you whether or not you're being successful,” Willis says. “But they can't tell you how people feel in their roles.”
The true challenge with AI and the workforce won’t just be how it mitigates bias, but how it expands the talent pool by identifying those who are often overlooked, according to the World Economic Forum report.
Today, most companies use AI in some form to automate their hiring practices. Those tools can carry their own biases if they're not trained properly, some experts warn.
Julie Cafley, the executive director from Catalyst Canada, says organizations deploying any type of AI program need to understand the data they're training on. She adds that they should work to reverse bias as part of the process to ensure proper representation of different groups.
Cafley shares an example of a colleague who joined Mila summer school —a Quebec research institute that focuses on AI and machine learning. “What was interesting is that she was the only person with an inclusion lens who was part of that initiative,” she says. “It was really entirely tech-focused individuals, and primarily men.”
