Entry-Level Jobs in the AI Era: What Changed for Graduates

Quick answer

Something real changed for graduates in 2025, and it is worth naming plainly rather than talking around. Australian graduate job postings fell roughly 15% across 2025, down around 35% from their 2023 peak, before stabilising in early 2026, according to the Australian Financial Review's reporting of Indeed Hiring Lab and Jobs and Skills Australia data. The mechanism is straightforward: the routine, repeatable tasks that used to fill a graduate's first year - first-draft research, basic financial models, assembling pitchbooks - are exactly what AI now does cheaply and fast. The door did not close. It got narrower, and the bar for getting through it rose toward AI familiarity and demonstrated capability. Teenagers who understand this early can prepare for the door that exists now, rather than the one that existed for their older siblings.

What actually changed in the graduate job market

Start with what the data actually shows, because the headline can sound worse than the reality. The AFR's reporting is specific: postings fell, then stabilised in early 2026, which is a correction, not a collapse. Jobs and Skills Australia's own 2025 analysis, Our Gen AI Transition - the first whole-of-labour-market look at generative AI in this country - found the technology augments far more work than it replaces across the economy as a whole. Graduates are not being locked out of work in general. They are being screened harder at the specific point where they used to enter it.

That distinction matters for how a family responds. Panic says "avoid this path altogether". The evidence says something narrower and more actionable: the entry rung changed shape, and preparation should change shape with it.

Why entry-level tasks got automated first

Graduate roles have always leaned on routine work, almost by design - juniors learn by doing the repeatable parts while senior staff handle judgement calls. That is precisely the profile AI absorbs fastest: predictable inputs, rules-based processing, a template to follow. First-draft modelling, structured research summaries, standard document assembly - the exact training-wheel tasks that used to justify hiring a graduate at all.

This is not a verdict on any particular industry or degree. It is a pattern that shows up anywhere a junior's job was mostly routine. Jobs and Skills Australia's broader finding, that AI lifts demand for problem-solving, communication and adaptability even as it absorbs routine tasks, is the flip side of the same coin. The tasks are moving, and the value is moving with them, toward the graduates who can do the parts that remain.

The rising bar: what employers actually want now

With the training-wheel tasks gone, employers are asking new graduates to arrive further along than before. The pattern is consistent across the evidence: PwC's 2025 Global AI Jobs Barometer found roles requiring AI skills carry a 56% wage premium, and grew 7.5% in postings even as overall postings fell. The World Economic Forum's Future of Jobs Report 2025 ranks analytical thinking as the single most important core skill employers want, with AI literacy the fastest-growing. Jobs and Skills Australia adds the human half of the picture: communication and teamwork now sit among the top graduate capabilities employers name.

Put together, the new bar is not "know how to use AI" in the shallow sense of having opened a chatbot. It is AI familiarity paired with the judgement to direct it, verify its output, and communicate the result - exactly the combination a transcript alone cannot show. For the wider context on how AI is changing Australian schools and the pathway toward this bar, see AI Education for Teenagers in Australia.

How teenagers can pre-empt the shift

The practical response starts years before graduation, not in the final semester of a degree.

1. Build AI familiarity early and often, not as a crash course before job applications. Directing AI tools well, and knowing where they fail, is a habit formed over years, not weeks.
2. Start a portfolio of real, finished work well before university - a researched project, a small build, evidence a student can frame a problem and finish it. Graduate Employability Starts at 15 sets out what that portfolio should contain and when to start it.
3. Treat communication as a core skill, not an extra. With communication and teamwork now among the top graduate capabilities employers name, practice presenting and explaining work matters as much as producing it.
4. Don't wait for university to teach this. Structured, project-based programs outside school hours can build the same judgement years earlier.

Common mistakes

• Reading the 15% fall as "there are no jobs". The market tightened and then stabilised; it did not disappear.
• Waiting until the final year of a degree to build AI skills. By then there is no runway left to build a real body of work.
• Assuming grades alone still clear the bar. Employers increasingly want evidence, not just a transcript.
• Avoiding a field because it is "getting automated". Most fields are being reshaped, not eliminated; the judgement-heavy parts of every field are growing.
• Treating AI fluency as a one-off course. Jobs and Skills Australia's evidence points to ongoing adaptability, not a certificate ticked off once.

The recommendation: take the 2025 graduate data seriously without treating it as a verdict on any single field. The door narrowed because the routine tasks that used to train juniors got automated first, and it will keep favouring graduates who arrive with demonstrated AI fluency, communication and a body of real work. Start building that years before the degree ends, not in the final semester. How a Student Can Build a Portfolio That Stands Out Before University and Is University Still Worth It in the AI Era? set out the practical next steps.