Expert notes importance of ‘creating psychological safety in the work environment for women to feel like they can use AI’
Training alone is failing to increase artificial intelligence (AI) adoption among women, a gap HR professionals will need to address as workplace AI use accelerates and retention risks tied to inequitable adoption grow, according to a recent report.
Currently, training increases AI adoption among men but has no measurable effect on adoption among women, notes The Prosperity Project, an organization focused on advancing women in the economy.
Julie Savard-Shaw, executive director of The Prosperity Project, says training programs vary by sector, ranging from basic prompt-writing instruction in fields such as agriculture and finance to advanced automation training in technology-focused companies. Regardless of sophistication, she said, training does not address the core barrier facing women.
"What's really needed is that AI training needs to be done alongside something that's really important, and that is creating psychological safety in the work environment for women to feel like they can use AI," she said in an interview with Canadian HR Reporter.
Business leaders say they are now confronting “AI fatigue” among staff who feel swamped, regardless of high levels of adoption – and that weaknesses of training programs, according to a previous report.
Perceived competence penalty
The Prosperity Project report found that when identical AI-assisted work is evaluated, perceived competence drops twice as steeply for women as it does for men. Savard-Shaw said the barrier stems from employer perception rather than women's own confidence or skill.
"It's actually the perception of the employer," she says, adding that similar double standards shape how traits such as assertiveness are judged differently by gender in Canadian workplaces.
The report frames this as evidence that AI use is socially evaluated through existing workplace and gender norms, rather than through skill alone. Where performance evaluation and promotion are at stake, the report states, avoiding tools that trigger negative perception can become a strategic choice rather than a skills issue.
Currently, 43 percent feel guilty using AI to produce work, a number that climbs to 56 percent among Gen Z employees, according to a previous report.
Exposure and data gaps
The Prosperity Project report cites Canadian labour market data showing 71 per cent of women work in occupations exposed to AI, compared with 49 per cent of men. It also found women are more likely to experience automation and increased monitoring, while men are more likely to benefit from AI augmentation tied to decision-making and strategic work.
Savard-Shaw says Canada still lacks comprehensive national data tracking how AI affects women across industries, regions and intersecting identities, with data on Black women, racialized women, gender-diverse professionals and workers with disabilities largely absent from the national picture. Existing analysis remains concentrated primarily in Quebec and Ontario, the report notes, with limited comparable national measurement.
"It's a really stressful environment to be in because you understand that there is an opportunity, but you also understand greatly the risk, and the risk is much higher for women," she says.
Recommendations for employers
The =report calls on organizations to strengthen AI governance, design AI systems that expand rather than compress work, and create environments where employees can experiment with AI tools without disproportionate professional risk. Savard-Shaw recommends employers adopt existing frameworks, such as one developed by the Ontario Human Rights Commission, rather than building new assessment tools independently.
She also urged organizations to set clear, consistently applied AI-use expectations for all employees regardless of gender, noting that Canada has made international commitments around responsible AI, though governance across the private sector remains largely voluntary.
"Really setting those standards and being transparent about what is and is not acceptable and then having the same apply to everyone, no matter the gender,” is important, she says.
“So if you're a man that uses AI, you're treated and your performance is treated the exact same way as if you're a woman using AI."
Without taking these steps, there's the risk of a homogeneous workforce, Savard-Shaw says, "and that ultimately will not be beneficial for your profits at all."
