Will AI Take My Child's Job? An Honest Answer

Quick answer

Almost certainly not their whole job, but quite possibly some of the tasks inside it. That distinction is the entire answer, and it matters more than it sounds. Jobs and Skills Australia's 2025 research, the first whole-of-labour-market Australian analysis of generative AI's impact, found the technology augments far more work than it replaces, reshaping roles task by task rather than deleting them wholesale. At the same time, graduate job postings did fall around 15% in 2025 before stabilising, concentrated in the routine, mechanical tasks that used to fill a graduate's first year. Both things are true at once: jobs are not disappearing en masse, and the entry rung of many careers is genuinely changing shape. The honest answer to "will AI take my child's job" is: not the job, but possibly the version of it you're picturing - and preparation for that shift beats trying to predict exactly which job title survives.

The question every parent asks, and why it's the wrong shape

"Will AI take my child's job" is the question almost every parent eventually asks, usually somewhere between a news headline and a quiet moment of genuine worry. It is a completely reasonable thing to lie awake thinking about. It is also, in its most common form, the wrong shape of question - because jobs are not single, solid things that either survive or vanish. They are bundles of dozens of individual tasks, and AI's actual effect is on the tasks, not the bundle as a whole.

A lawyer's job is not one thing; it is drafting, researching, advising, negotiating, arguing, building trust with a client under stress. AI is currently good at some of those tasks and genuinely poor at others. The same is true of a nurse's job, a teacher's job, an accountant's job. Asking "will this job survive" skips past the far more answerable and far more useful question: which of the tasks inside it will change, and what does my child need to be good at once they do?

What the evidence actually shows: two findings, not a contradiction

Two pieces of Australian evidence get quoted at parents, and they can sound like they disagree. They do not.

Jobs and Skills Australia's 2025 analysis, Our Gen AI Transition, is the reassuring finding: generative AI augments far more work than it replaces across the whole economy, and lifts demand for problem-solving, communication and adaptability as it spreads. The Australian Financial Review, citing Indeed Hiring Lab and Jobs and Skills Australia data, is the sobering one: graduate job postings fell around 15% in 2025, roughly 35% below their 2023 peak, before stabilising in early 2026 - concentrated in roles where a graduate's first-year work was heavy on routine, mechanical tasks like financial modelling and first-draft research.

Read together, not against each other, the picture is coherent. Across the whole labour market, AI is augmenting work and raising demand for human skills. At the entry rung specifically, the kind of task that used to fill a junior person's first year or two is exactly the kind AI now does cheaply, so that rung is genuinely narrower and the bar for standing out has risen. Neither finding cancels the other. The honest synthesis is: the ladder is not gone, but its bottom rung has changed shape, and standing on it now takes more than a degree used to require.

Task-level, not job-level: what this actually means for your child

The World Economic Forum's Future of Jobs Report 2025 adds the scale of the shift: it expects 39% of core work skills to change by 2030. That is a genuinely large number, and it is also not a number about jobs disappearing - it is a number about what skills those jobs will need. The two are different claims, and the second one is the actionable one.

Preparation beats prediction: what actually helps

You cannot forecast the labour market a decade out with any real confidence, and any guide that claims otherwise is overselling. What you can do is prepare your child with skills that hold value across a wide range of futures, which is a strictly better strategy than betting on one predicted outcome.

1. Build analytical thinking deliberately. The World Economic Forum ranks it the single most important core skill in the workforce, and it is built through genuine struggle with hard problems, not shortcuts around them.
2. Add real AI literacy, not just AI use. Understanding how these tools work, where they fail, and how to direct and verify them is now table stakes, not a specialist skill - our guide to AI education for teenagers in Australia covers how that literacy is built at home and at school.
3. Treat communication as core, not optional. Jobs and Skills Australia now ranks communication and teamwork among the top graduate capabilities employers want - it is not a soft add-on to technical skill, it is co-equal with it.
4. Get your child building real things, early. A portfolio of finished, real projects is what stands out against the routine, mechanical work AI now absorbs cheaply at entry level - the practical detail is covered in how students build a portfolio before university.
5. Resist the urge to pick a "safe" job title. No title is safe on a multi-decade horizon. Skills that transfer across many possible futures are the actual insurance policy.

Common mistakes parents make with this fear

• Reading a single scary headline as the whole picture. The AFR's graduate-posting decline and the Jobs and Skills Australia augmentation finding both need to be held together, not cherry-picked.
• Trying to pick the one "AI-proof" career. This bets everything on predicting an unpredictable decade. Broad, durable skills are the more reliable strategy, covered further in future-proof careers for teenagers.
• Treating the anxiety as something to hide from a teenager. Naming it plainly and calmly, with real evidence rather than headlines, usually reduces a teenager's own worry rather than adding to it.
• Waiting to "see how it plays out" before doing anything. The skills worth building - analytical thinking, AI fluency, communication - compound over years. Waiting is not neutral; it is a choice to start later.

The recommendation: stop asking whether a specific job title will still exist, and start asking whether your child will be able to do the parts of future work that AI genuinely cannot. Hold both pieces of Australian evidence together - augmentation across the economy, a narrower and more demanding entry rung - and prepare accordingly: analytical thinking, real AI literacy, strong communication, and evidence of things actually built. That preparation travels across whichever specific jobs exist in fifteen years. A prediction about one job title does not.