The good, the bad and the biased: How AI is changing hiring
AI is reshaping how companies hire, train and manage workers. But according to researchers at TMU's Diversity Institute (DI), how businesses adopt these tools matters just as much as whether they do.
In early June, Canada’s federal government launched AI for All (external link) , a national strategy with the goal of boosting AI adoption among Canadian businesses from about 12 per cent today to 60 per cent by 2034. Small and medium-sized enterprises (SMEs) are a top priority – they make up more than 99 per cent of Canadian employer businesses and employ nearly two thirds of private sector workers, yet continue to face persistent labour and skills shortages that constrain growth.
Research from the Diversity Institute underscores the urgency. Their (PDF file) December 2025 survey found that improving awareness of AI tools and their practical applications, supported by targeted funding, could significantly accelerate adoption among small and medium-sized enterprises (SMEs).
But Wendy Cukier, professor of entrepreneurship and innovation and founder of the Diversity Institute, says “AI is a double edged sword – it can improve processes and level the playing field or deepen disadvantages.”
DI has been leading research on AI trends and responsible AI adoption with a focus on organizational change.
“If you don't have inclusive hiring processes to begin with, AI will just replicate bias,” she says.
Wendy Cukier, founder and academic director of the Diversity Institute, and academic research director of the Future Skills Centre, sees the transformational potential of AI, stressing the importance of creating awareness, access and support for human-centred adoption by businesses.
Right now, large companies are using AI in hiring but most SMEs are not. The Diversity Institute/Memorial University survey conducted between April and August 2025 found that 28 per cent of organizations struggled to fill job vacancies, with nearly one-third struggling to attract suitable candidates, yet only 4 per cent of employers reported using technologies like job platforms or resume screening software.
This gap represents significant untapped potential for AI to modernize hiring and expand access to talent. Yet, the risks are real. One study found that bias in AI hiring tools can directly influence decisions (external link) , with managers more likely to mirror that bias than correct it.
Research from the Diversity Institute has shown:
- There is a gap between Canada’s leadership in the development of AI and its adoption of tools.
- How AI is defined has an impact on our understanding of its adoption. For example, a December 2025 survey with Memorial University showed clear opportunities for strengthening “use cases” and support for SMEs to accelerate adoption.
- The impacts of AI are not seen uniformly but will affect women, Indigenous peoples and other equity deserving groups in different ways that could exacerbate or ameliorate job loss, bias and the digital divide.
- Advancing responsible adoption, particularly among SMEs, requires capacity building and advisory supports, as well as access to tools and technology. DI has proposed an AI competency framework and playbook as well as AI Youth Corps to encourage adoption, create opportunities and develop talent.
Built-in bias
The federal AI strategy aims to empower Canadians to participate in and benefit from AI. This means supporting better understanding and access that enables Canadians to shape how AI is used. Photo credit: Gender Spectrum Collection
AI hiring tools learn from historical data. That means they can embed biases that reinforce past hiring patterns, like hiring more men over women, favouring certain universities or focusing on “Canadian” experience. Systems may also penalize applicants with employment gaps (for things like parental leave or disability leave).
Bias can also enter earlier, at the job description stage.
“Often there are qualifications that are included in job descriptions that are really there as sorting mechanisms, they're not actually needed to perform the job.” — Wendy Cukier, founder and academic director, Diversity Institute
One common example: requiring post-secondary education when it isn’t essential to the role.
“As soon as you put in post-secondary education and don't add ‘or equivalent’, you're going to exclude a large segment of the population,” says Cukier. That includes, disproportionately, Indigenous Peoples who are far less likely to be university graduates but may have the competencies needed to do the job.”
“It’s important to question assumptions and build safeguards into hiring systems at every stage.” — Wendy Cukier, professor of entrepreneurship and innovation at the Ted Rogers School of Management
Powering-up HR
Cukier says responsible AI in HR comes down to six principles: fairness, reliability, privacy, inclusiveness, accountability and transparency. In practice, that means keeping humans in the decision-making loop.
“AI systems can help screen candidates or generate shortlists, but final decisions must always remain with people,” she says.
When implemented thoughtfully, AI can expand access to more diverse talent pools and improve workforce planning by matching skills to business needs, identifying training gaps and to promote equitable advancement opportunities.
AI for All
Not all businesses are using AI the same way. Early findings from the Diversity Institute suggest large organizations, including banks, tech firms and even the federal government, are often using AI to cut costs and reduce their workforce. In contrast, smaller businesses tend to use it to grow, expand capacity and address labour shortages.
The challenge for SMEs is access: knowing which tools exist, being able to afford them and having the expertise to use them well. Despite the rise in demand for AI skills, DI’s research shows that fewer than one in ten SMEs report access to formal AI training. Also, employers often don’t have the frameworks in place for their workers. Nearly one-half of employees who are using AI tools at work received no training, and over one-third had minimal employer guidance which in turn creates serious security risks.
Canada's AI for All strategy aims to address that gap – and Cukier sees a role for young talent to support the development of AI skills in young Canadians while helping SMEs implement practical tools safely and effectively.
The opportunities are significant. With the right supports in place, broader AI adoption could drive both innovation and economic growth, and create a stronger economy for all Canadians.