Does AI give the next generation a head start, or shortcut the learning that builds real skills? 57% of Gen Z already use AI at work. Most can't evaluate what it gets wrong. Both sides, steel-manned, with every claim traced back to a named source.
Eight parts, one cover. Same frame every issue in The Debate; only the words and the two filled grid cells change.
The assumption that young people are already equipped for an AI-saturated workplace because they grew up online is not just wrong — it is harmful. It removes the urgency to teach them anything, and it leaves them exposed to a skills gap they cannot see because they assume they have already crossed it.
EY research on young people who rated themselves "very knowledgeable" about AI found that the same group scored poorly when asked to write effective prompts or identify where an AI output had gone wrong. Being fluent in social media, comfortable managing multiple devices, and quick to adopt new apps does not transfer automatically to knowing how to interrogate a language model, evaluate its limitations, or use it as a precision tool rather than a shortcut. Those are distinct skills, and they are learnable — but learning them requires being told they need learning.
The flip side is real too. 57% of Gen Z professionals already use generative AI at work — higher than any other generation. People who are already using it, even imperfectly, are accumulating experience and building instincts that compound over time. The opportunity gap between young people who invest in AI literacy now and those who do not is already opening. The question is whether it widens into something structural.
Each side, put as well as its own advocates would put it. Sources named. No figure here is offered as settled fact; each is a claim by a named author, linked in full at the end.
EY research: even young people who rated themselves "very knowledgeable" about AI scored poorly when writing prompts for or evaluating shortfalls of AI tools. Digital native does not mean AI-literate. The assumption that it does is actively harmful — it removes the urgency to teach.
Workers aged 22–25 in AI-exposed roles have already seen a 16% drop in employment. The youngest and least established workers bear the earliest, sharpest impact of AI-driven suppressed hiring — before they have accumulated the experience that protects older workers.
Demand for digital skills is accelerating far faster than global supply. The skills gap is widening at precisely the moment young people are entering the workforce — the timing is the worst it could be for the cohort now in school.
— WEF — Bridging the digital skills crisis (Dec 2025)
92% of US jobs now require digital literacy, but the education system is still building that curriculum. Many students leave school with neither the AI skills nor the critical-thinking framework needed to evaluate AI outputs — two separate gaps that are often conflated.
— EDUCAUSE 2026 Top 10 — Technology literacy for the future workforce
57% of Gen Z professionals already use generative AI at work — higher than any other generation — for data analysis, creative work, content creation, and project management. They are not waiting for permission or for training to arrive. They are learning by doing.
Gen Z grew up managing digital information overload, evaluating social-media sources, and adapting to rapid platform changes. Constant re-evaluation and comfort with multiple parallel tools are directly relevant to working with AI systems that change frequently and require ongoing calibration.
— Deloitte 2025 Gen Z and Millennial Survey
Young people who invest in AI literacy now face a widening opportunity gap compared to peers who don't. Early movers build skills that compound over careers measured in decades — a small head start in 2025 could translate into a substantial advantage by 2035.
— WEF Future of Jobs Report 2025
The skills most needed — critical thinking, judgment about AI outputs, ethical evaluation of automated decisions — are learnable. They are not locked to any particular generation, and they are not exotic. They are the skills that education systems already know how to teach, applied to a new context.
House rule: any percentage about employment or skill levels is attributed to its author and study. We reference claims; we do not manufacture certainty about contested generational data.
The EY overconfidence finding is important not because it proves young people are behind, but because it identifies a specific mechanism by which they stay behind: the belief that no work is required. An assumption of competence is the one thing guaranteed to prevent competence from developing.
Using TikTok and knowing how to evaluate an AI output are entirely different skills. One is a consumption habit shaped by algorithmic nudges; the other is a disciplined practice of forming a question, assessing the response, testing the limits, and checking the output against independent evidence. Young people are not worse at the second than anyone else. They are often better-positioned to learn it quickly. What they are missing is the signal that it matters, and the structured context in which to practise it.
The good news is that the skills most needed are not exotic. Critical thinking about source quality, recognition of plausible-but-wrong outputs, and judgment about when to trust a tool and when to verify independently are all things education systems know how to teach. They existed before AI and they apply directly to it. The curriculum challenge is not starting from scratch; it is adaptation and prioritisation — getting these skills into classrooms and workplace training fast enough to matter for the people who need them now, not in five years.
Whether prompt engineering specifically is a durable skill or will be obsoleted by better models in two years is a live debate. The answer determines what exactly to teach — and when. The more defensible bet is to teach the meta-skill: how to work with tools that are more capable than you in some dimensions and less reliable than they appear in others.
Three actors are in a position to close the gap. Currently none of them is moving fast enough.
The EDUCAUSE data shows the curriculum is behind the labour market. The practical fix is not a new subject called "AI" — it is integrating critical evaluation of AI outputs into existing subjects across the board. History classes can evaluate AI-generated narratives. Science classes can test AI on empirical questions and document where it fails. English classes can distinguish AI-assisted from human-authored work, and discuss why that matters. The goal is not AI specialists; it is people who can use a powerful tool without being fooled by it.
The 57% of Gen Z already using AI at work are largely learning by trial and error without structure. Employers who build formal AI literacy into onboarding — not as a compliance exercise but as practical tool calibration — get better outcomes and accelerate the transition for the least-experienced workers who need it most. This is in employers' direct interest; it is also currently uncommon.
The opportunity gap is not evenly distributed. Young people at well-resourced schools and workplaces will get structured AI literacy whether or not it is explicitly on the curriculum, because they are surrounded by people who already have it. Young people in under-resourced environments will not. If no one is paying for the upskilling — not the school, not the employer, not the state — the gap compounds into something structural. Access to AI tools without the skills to use them critically is not a head start. It is a different kind of risk.
Every claim above, traced to its author. Read the originals; argue with them, not with us.
The Debate — No. 07 · AI and Young People · Craig Stanley Studio · CS·PRESS · Undated by design, refreshed as the argument moves · One red, used once · All colour pairings meet WCAG 2.1 AA.
Nine debates, each titled and never dated, each refreshed in rotation as the argument moves. Plain English. Both sides put at their strongest. Every claim about the world attributed to its author with a link. No hype, no banned words, no manufactured certainty.
AI and Creativity · AI and the Environment · AI and Ethics · AI and Power · AI and Truth · AI and Work · AI as a Force for Good · AI in Education.