AI could screen out newcomers’ job applications, study warns
AI could screen out newcomers’ job applications, study warns
https://toronto.citynews.ca/2026/07/04/ai-could-screen-out-newcomers-job-applications-study-warns/
Publish Date: 2026-07-04 08:13:00
Source Domain: toronto.citynews.ca
As more employers use artificial intelligence (AI) to screen job applications, there is growing concern that AI could amplify biases and limit newcomers’ chances of getting hired.
“That is the risk of AI,” explained Kristin Schwartz, senior manager at CultureLink. “If there is a pattern of favouring candidates for particular Canadian universities, for example, AI will quietly learn that that is what a good candidate looks like.”
Schwartz told OMNI News that learning about how employers are using AI can be helpful, and that while there are systemic barriers, the clients at her settlement agency shouldn’t give up hope.
“Everybody is trying to come to grips with an AI which is screening out people who don’t fit the mould more efficiently than ever before,” she said.
Concerns regarding the use of AI and a lack of transparency and fairness in the hiring process are not new, but researchers warn that immigrants could be uniquely impacted.
Rupa Banerjee, Canada Research Chair in Economic Inclusion, Employment and Entrepreneurship, noted that “when you layer individual human bias on top of algorithmic bias, it’s not surprising that you are going to see some people be disproportionately disadvantaged over others.”
Banerjee’s research at Toronto Metropolitan University is looking at what happens when AI decides who gets access to job opportunities.
What they are hearing from both employers and job seekers is that the lack of humanity in the process is fuelling mistrust, and that newcomers are feeling disappointed and disillusioned.
“It seems to actually activate more bias,” she told OMNI News, adding that one of the employers she spoke to said they were very worried about even bringing in an immigrant for an interview because now they “don’t know if maybe they just use ChatGPT and they can’t speak English at all, and it’s much more risky than it would have been in the past, when we could just screen out…
