Sixty-seven million young people aged 15 to 24 are out of work worldwide, and the entry-level clerical and administrative roles that have historically served as career on-ramps are the occupations AI is reshaping fastest. The International Labour Organisation's 2025 youth employment report arrives at a moment when Statistics Canada's Labour Force Survey is already tracking a Northern America youth unemployment rate that climbed from 8.3% in 2023 to 9.8% in 2025, placing Canada squarely inside the pattern the ILO describes. For a 22-year-old lining up applications this autumn, and for an HR lead designing next year's graduate intake, the numbers come from the same source. The implications pull in different directions.

What the ILO actually reported

Global youth unemployment reached 12.4% in 2025, with 67 million people aged 15 to 24 out of work and more than 257 million young people classified as not in employment, education, or training (NEET). Youth unemployment worsened in 8 of the world's 11 subregions between 2023 and 2025. In Northern America, which includes Canada, the rate climbed from 8.3% to 9.8% over that same period. Across Northern, Southern and Western Europe the rate held at 15%, and 20 of 29 countries in that grouping reported worsening job prospects for young workers.

The ILO report characterizes the moment with unusual directness: the tentative, and short-lived, recovery in youth labour markets that followed the COVID-19 pandemic has stalled. The NEET figure carries its own warning. The report notes that with many more young people implicated and with youth disengagement harder to redress, the increasing trend in the youth NEET rate is especially worrisome. For Canada, Employment and Social Development Canada (ESDC) tracks NEET trends through its own Labour Market Information systems, and the ILO's headline direction aligns with that domestic picture.

The AI angle: it is the entry rung that is exposed

The ILO estimates that 6.1% of jobs held by workers aged 15 to 29 are in occupations highly exposed to AI-driven change. Those exposed roles are largely middle-skilled clerical, administrative and service positions - the same roles that have traditionally served as the first rung on the career ladder. In Canada, that category maps onto the kinds of entry-level positions that have historically bridged a college diploma or undergraduate degree into stable employment at institutions from Bay Street banks to federal departments and provincial ministries.

The ILO's stress-test scenario is instructive: if only 10% of the jobs exposed to AI were to disappear in full, it would translate into 5.6 million employed youth - largely in high-income countries - facing unemployment, changing jobs, or exiting the labour force. That is not a forecast; it is a modelled pressure point. Canada, as a high-income economy with a significant share of white-collar entry-level work concentrated in finance, insurance, real estate and federal public service, sits in the portion of the risk distribution the scenario describes most precisely.

Gilbert Houngbo, ILO Director-General, framed the consequences in terms of social cohesion: a generation that cannot find decent work cannot build its future with confidence, and when young people are locked out of quality employment, countries lose talent, productivity and social cohesion. Sukti Dasgupta, ILO Director of the Employment, Skills and Sustainable Enterprises Department, added a note of caution about pace: there is increasing noise around AI, and while the direct impact on jobs is still unclear, we must not be complacent and underestimate the risks.

For young workers and candidates

If you are in your early twenties and trying to land the first job that actually builds a CV, the market you are entering is structurally different from the one that greeted graduates five years ago. The ILO numbers are not an invitation to panic. They are a prompt to reposition. Here is what that looks like in a Canadian context.

Assume the entry-level ladder has fewer rungs, and look for routes that build differently. The classic path of landing any admin role, proving yourself on routine tasks, and climbing from there is thinning in precisely the segment AI is reshaping fastest. Look for entry roles that combine human judgement with a technical dimension: junior analyst positions where you are reviewing and correcting model outputs, client-facing work where relationship skills are the core competency, or apprenticeship routes that embed structured learning into paid work. Canada's Red Seal Program and provincial apprenticeship frameworks - administered jointly by ESDC and provincial bodies such as Ontario's Ministry of Labour, Immigration, Training and Skills Development and Quebec's Commission de la construction du Quebec (CCQ) - offer credentialed pathways that remain largely hands-on and AI-resistant by their nature.

Treat AI as a tool you can speak to confidently in an interview, not a threat to sidestep. Employers hiring in 2026 increasingly expect junior candidates to have practical experience with generative tools. If you can describe how you used AI to draft or sort something, then reviewed it, caught an error, and improved the output, you are demonstrating exactly the layered judgement that the roles surviving displacement reward.

Widen the search geographically and by sector. The ILO data shows uneven exposure across regions and industries. Sectors less exposed to AI displacement - skilled trades, healthcare, personal care, and technical roles with a hands-on component - are actively hiring across most Canadian provinces. If your applications have been concentrated in the office-based roles your credential ostensibly pointed at, broadening the funnel is a rational response to a narrowing pipeline.

Invest in the skills the ILO explicitly flags. Dasgupta's call to step up investment in skills, lifelong learning and social protection so that young people can adapt to change and seize new opportunities - and to ensure that technological progress, including AI, works for young people rather than against them - translates, at the individual level, into short and verifiable skill signals. Data literacy, structured communication, a working familiarity with at least one AI toolset, and one domain-specific competency anchoring your CV to something a language model cannot replicate - accounting basics, a coding stack, a technical qualification, or French-language proficiency in Quebec - all sharpen your positioning in a compressed entry market.

For HR leaders and employers

The part of the ILO report that creates the most strategic exposure for HR teams is not the headline unemployment number. It is the second-order effect: if the roles you have traditionally used to seed the base of your talent pipeline are the same roles AI is compressing, the pipeline is quietly fracturing, and that fracture may not surface in your workforce metrics until three to five years out, when the middle-management bench you assumed you were building is thinner than expected.

Audit what your entry-level roles actually contain. A significant share of junior job descriptions written between 2020 and 2023 are, in 2026, describing work that is now partly automated. That does not necessarily mean the role should be eliminated. It means the role likely needs rescoping: identifying the human-judgement layer, the relationship work, and the learning-by-doing that was previously embedded in routine tasks. If a junior team member spends the majority of their time on work an AI now completes in seconds, the role warrants redesign before it becomes a costly mismatch between salary and function. Chartered Professionals in Human Resources (CPHR Canada), through its provincial associations including HRPA in Ontario and CPHR BC/Yukon, has published competency frameworks that can help guide that rescoping toward higher-judgement role profiles.

Rebuild apprenticeship and structured learning routes. Where the traditional entry rung is compressing, apprenticeship schemes and rotational graduate programmes become more durable pipeline investments. Canada's registered apprenticeship system, supported through ESDC's Apprenticeship Incentive Grant and Apprenticeship Completion Grant, provides a federal funding scaffold for employers in designated Red Seal trades. For non-trades sectors, co-op partnerships with colleges and universities, formal rotational schemes, and paid internship agreements with post-secondary institutions offer comparable structure. Employers who invest in scaling these arrangements over the next 18 months are better positioned to have a functional junior bench in 2029.

Treat reskilling as a retention lever, not simply a training budget line. The ILO's modelled scenario of 5.6 million displaced youth concentrated in high-income economies lands hardest on precisely the workforce segment most Canadian employers depend on for early-career pipeline. Internal mobility programmes that move people from AI-exposed roles into adjacent, higher-judgement work are typically less expensive than running a full recruitment cycle two years down the line, and they signal to the remaining workforce that AI adoption is a workforce strategy with a human dividend, not only a cost story. The Conference Board of Canada's compensation and workforce research tracks internal mobility as a retention indicator worth monitoring alongside standard attrition metrics.

Widen how you screen. If AI has compressed the productivity gap between candidates on routine tasks, then traditional filters such as degree classification, institution tier, and target-school pipelines are less predictive than they were a decade ago. The variance that matters now sits in judgement quality, intellectual curiosity, and the demonstrated ability to work with AI tools critically rather than passively. Structured behavioural interview questions probing those attributes will out-predict a first-round CV screen calibrated to criteria that predate the current AI cycle. Under the Canadian Human Rights Act and applicable provincial human rights codes, screening criteria should also be reviewed for adverse effect on protected groups, given that credential-based filters can disproportionately disadvantage equity-seeking candidates.

The shared thread: the pipeline is a joint problem

The ILO called for a human-centred approach to AI governance, arguing that innovation must create opportunities for young people rather than close them off. The report's own framing is worth reading in full: rebuilding confidence among today's youth will require renewed investment in decent work, stronger labour market institutions, and policies capable of helping young people navigate an increasingly uncertain future while ensuring that technological progress and economic transformation work for, rather than against, the next generation.

Gilbert Houngbo's closing framing holds for both sides of the Canadian interview table: creating decent jobs for young people is not just a social imperative but one of the smartest investments a country can make. That logic scales to the employer level. Creating a genuine entry-level position, one where a 23-year-old learns something substantive and contributes something real, is a particularly high-value investment precisely because the market is producing fewer of them. The Canadian Labour Congress (CLC), which represents approximately three million union members nationally, has consistently flagged youth underemployment as a structural concern alongside wage stagnation, and the ILO data gives that concern a sharper empirical edge in 2025.

For candidates, the corollary is the same in reverse. Roles that look like slow starts - apprenticeship placements, junior positions with genuine training content, technical trainee schemes in the trades or in regulated industries such as banking and telecom - carry the strongest structural floor in an AI-reshaped labour market. The 67 million figure will not shift on its own. It moves when the design of entry-level work moves. That is a decision taken on both sides of the interview table, and in Canada, it is one that ESDC policy, CPHR Canada practice standards, and the Red Seal apprenticeship system are all positioned to support if employers and candidates choose to use them.