The Entry-Level Job Is Dying. AI Skills Could Save It.

For more than a decade, I’ve worked in employment services, helping young people push past the barriers that keep them out of good jobs. For years those obstacles were familiar: the experience paradox—you need experience to get a job, but you need a job to get experience—and the absence of networks to open doors. Now, AI has added a new layer of difficulty.

AI is rapidly reshaping the job market in Canada, and young people are bearing the brunt of this seismic shift. Entry-level roles that once served as footholds into the labour force, like customer service, administration or retail, are disappearing. Bloomberg reports that AI could replace more than half the tasks done by market research analysts and sales reps, compared to just nine per cent and 21 per cent of their managers’ tasks. And seasonal and summer job markets in Canada, historically a source of entry-level work, are at their worst in two decades.

At the same time, employers increasingly expect job candidates to demonstrate fluency with AI tools. More and more job postings mention AI expertise. But these are skills that many young people haven’t had the opportunity to learn. Post-secondary institutions aren’t keeping pace.

I see AI’s impact on the job market in real time through my work at Skills for Change, a Toronto-based nonprofit that provides employment training, mentorship and skills development for immigrants, racialized communities and youth under the age of 30. Since the release of ChatGPT in late 2022, more young people have been coming through our doors feeling overwhelmed and anxious about their future. Many of them tell us the jobs they assumed would be available, the roles they’d been studying and training for, have either disappeared or changed beyond recognition. On top of that, they’re grappling with Canada’s rapidly rising cost of living. Many people from lower-income communities are worried that AI could permanently block their path to economic mobility. Their concerns are reflected in the data: in the first quarter of 2025, Canada’s youth unemployment rate for graduates aged 15 to 24 hit 11.2 per cent—the highest it’s been since the mid-1990s, excluding the pandemic.

AI is here to stay. The good news is that many AI-related roles are skill-based rather than degree-based. If young people can access the right training, they can have a shot at breaking into fast-growing fields—without spending years in a classroom. That’s why, this past spring, we launched an AI upskilling program at Skills for Change. We worked with Google to help young people not just adapt to AI, but use it to their advantage.

In April, we rolled out our first pilot. Delivered over three weeks, the pilot followed a hybrid model combining instructor-led, in-person sessions with independent virtual learning. It integrated Google’s AI Essentials course, a beginner-level program that teaches practical uses of AI tools in everyday work settings. For example, students deconstructed job postings, identifying the specific skills and language employers were looking for, then used AI to craft resumés and cover letters that spoke directly to those needs.

The first group of participants, youths aged 18 and up, came to us with a range of career interests: some were interested in software development, while others were looking into hospitality, customer service or administrative roles. A few were still trying to figure out which direction to take. But all of them wanted to understand how AI could help shape and improve their career prospects.

The feedback from our first pilot was overwhelmingly positive. Participants came away feeling more confident about navigating the job market with AI as a support tool. However, some of our other younger clients expressed discomfort with our focus on AI upskilling. While Canadian youth are the quickest to adopt AI, many are far slower to trust it. They don’t understand its capabilities, and just as importantly, its limitations. That’s why our programming is about questioning AI, too. We taught participants about algorithmic bias, misinformation and the importance of human oversight, so that participants could use AI tools thoughtfully across a variety of real-world scenarios. The goal wasn’t rote learning. It was to help them build the confidence to think for themselves as AI continues to evolve.

Some of our clients worry that using AI to draft a cover letter feels like cheating or presenting a false version of themselves. But AI doesn’t replace their voice; it helps refine it. These tools are meant to augment what they already know and do, not fake those things. Our goal isn’t to replace anyone with AI—it’s to support people in using it effectively and ethically. 

We’re now embedding this AI training within all our programs, like settlement services and career planning. We’ve also created sector-specific bootcamps focused on fields like data analytics, accounting and finance. Over the next two years, we aim to provide AI training to 20,000 people. We’ll also track more detailed outcomes like employment results, self-reported comfort with AI tools and increases in AI competency scores. We’ll survey whether participants are actively using tools like ChatGPT, Gemini or Claude to build their resumes or explore new career options. 

Canada has a national AI strategy that emphasizes support in AI research and innovation. But it barely addresses the workers who will be directly affected by AI. There’s federally supported AI training but it’s concentrated on those in government and research. It largely overlooks youth outside of post-secondary institutions, even though they’re just as vulnerable to disruption and just as capable of contributing to AI-powered work. An AI upskilling program allows us to deliver accessible training to young people who might otherwise be left behind. But this kind of work can’t rely on the goodwill of a few corporations or be confined to a handful of non-profits. If we’re serious about preparing the next generation for an AI-driven economy, we need sustained, large-scale investment at the federal level. 

If we fail to act, we risk falling behind other countries that are already integrating AI training into every facet of workforce development. Countries like Estonia and Singapore are even incorporating AI education for children in primary and secondary schools. Without urgent upskilling, we risk widespread job displacement, particularly in sectors like customer service and administration. The consequences would be severe: increased structural unemployment, reduced innovation and a widening productivity gap that would undermine Canada’s ability to compete internationally. The global AI economy is projected to hit more than US$15 trillion by 2030; if we don’t move fast, those dollars will flow elsewhere, leaving Canada behind both economically and socially.

AI upskilling programs should be implemented across the full spectrum of public education and workforce development, from K–12 schools to skilled trades programs, libraries, community centres, settlement services and employment training non-profits. This is especially important for youth in remote geographic areas or from racialized or low-income backgrounds, because they’re the ones who are statistically most at risk of being left out of the evolving job market. This isn’t just about learning a novel tool. It’s about ensuring that newcomers to the job market can write stronger resumés, prepare more effectively for interviews and enter the workforce with skills that employers now expect.

—As told to Ali Amad

Surranna Sandy is the CEO of Skills for Change, a Toronto-based nonprofit that helps immigrants and underserved groups find jobs, learn languages and settle in Canada.