Jobs AI Cannot Replace in Australia: The Human-Advantage Roles Set to Grow Through 2030 product guide
Now I have comprehensive, authoritative data to write the article. Let me compile the verified final article.
The Counternarrative the Headlines Miss
Every week, another story warns that AI is coming for Australian jobs. Rarely does the coverage answer the more useful question: which jobs is AI structurally incapable of taking? The distinction matters enormously. Workers in the right occupations are not simply "safe" from automation — they are entering a labour market where demand will outpace supply through 2030 and beyond, partly because AI is reshaping the economy around them.
This article identifies those occupations, explains the structural mechanisms that make them resistant to automation, and grounds every claim in the most current Australian evidence available. If you are weighing a career decision in an AI-disrupted market, this is the data you need. (For the full picture of which roles are most exposed, see our guide on Which Australian Jobs Are Most at Risk from AI? A Role-by-Role Breakdown.)
Why Some Jobs Resist AI: The Three Structural Barriers
Before examining specific occupations, it is worth understanding why certain jobs resist automation. The answer is not simply that they are "human" — it is that they involve specific categories of work that current AI systems cannot replicate cost-effectively, or at all.
The International Labour Organization has produced exposure indices for more than 400 occupations, indicating the extent to which human input to each occupation will be displaced or augmented by AI. At one extreme, the most exposed occupation is data entry clerk, for which the ILO estimates 70% of the tasks currently done by humans could be done or improved by AI. At the other extreme, bricklayers and dental assistants are among jobs classified as "not exposed."
Three structural barriers explain this divergence:
1. Physical Dexterity in Unstructured Environments
AI excels in structured, predictable digital environments. It fails in chaotic, physical ones. A construction site changes every hour — ground conditions shift, materials arrive late, subcontractors make unexpected decisions, and safety hazards appear without warning. Robotics capable of navigating this variability at commercial scale remain decades away from widespread deployment. For construction, routine and repetitive tasks such as material handling and basic measurements will increasingly be supported by machines and digital tools, while humans focus on complex, adaptive, and safety-critical tasks. The construction industry should expect increased "bifurcation" of its workforce, with tech-enabled workers harnessing digital tools while still requiring large numbers of on-site workers.
2. Emotional Intelligence and Relational Care
AI can simulate empathy in text. It cannot hold a patient's hand, read the non-verbal distress of an elderly person with dementia, or make the split-second clinical judgment that distinguishes a nurse from a chatbot. Nursing, allied health, and individual support rely on empathy, clinical judgment, and communication — all human-first skills. These are not soft advantages — they are hard technical requirements that AI cannot currently meet.
3. Domain Expertise Applied to Complex, Adaptive Problems
CSIRO's Data61 research group notes that representation of domain knowledge is challenging because knowledge in the form of human expertise is large, dynamic, and non-uniform. Workers who combine deep contextual knowledge with the ability to make judgment calls in novel situations — midwives assessing foetal distress, electricians diagnosing a fault in a non-standard installation, social workers reading a family situation — are performing tasks that remain beyond AI's reach.
The Victoria University 2050 Simulation: What the Modelling Shows
The most rigorous Australian-specific modelling of AI's occupational impact comes from researchers at Victoria University's Centre of Policy Studies. Using a model of the Australian economy built on ILO research, they simulated two future versions of Australia through to 2050: one in which businesses and government adopt AI extensively, and one in which there is no AI. Comparing these two futures helps understand what we might gain and lose from this new technology.
The headline finding is not a jobs apocalypse. Overall, the simulation paints a picture of a larger and better-resourced economy, showing that total employment won't change a lot, but employment in some occupations will be much larger or smaller than it would be in a non-AI future.
Critically, the simulation identifies a specific mechanism by which AI-resistant roles grow: many studies, including the Productivity Commission's interim report on AI, find AI will drive faster economic growth. In a faster-growing economy, more people will work as teachers, hairdressers, and carers, because AI isn't expected to be as useful in those roles. This faster-growing economy will also require more school buildings, hair salons, and care homes.
This is a crucial insight that most media coverage misses entirely. AI does not just displace workers — it generates economic growth that creates demand for human-intensive services. The Victoria University researchers also recommend that government can facilitate early, jobs-focused investment in industries less exposed to AI, particularly those that require lots of interpersonal input — for example, investment in a world with fewer business analysts and more hospitality workers should be targeted at hotels and hospitality venues, rather than office space.
Jobs and Skills Australia's Projections: The Occupations Set to Grow
Jobs and Skills Australia (JSA) — the nation's peak labour market authority — publishes the most authoritative forward projections of Australian occupational demand. JSA has commissioned Victoria University to produce employment projections using the Victoria University Employment Forecasting (VUEF) model. The 2025 projections are unambiguous about which occupational groups will grow fastest.
Total employment in Australia is projected to grow by 961,000 people (or 6.5%) over the next five years, and by nearly 2.0 million people (or 13.3%) over the next ten years, reaching 16.6 million employed people by May 2035.
Within this overall growth, the standout sectors are overwhelmingly human-centred:
While growth across industries is broad-based, the greatest growth by far is expected in Health Care and Social Assistance, with its share of total employment projected to increase from 15.2% in 2023 to 16.7% in 2033.
Community and Personal Service Workers are projected to grow by 122,000 persons or 7.1% to 2030 and by 236,100 people or 13.7% to 2035. The fastest-growing specific roles within this group are striking: the fastest growth is projected to occur in Dental Assistants (26.8%), Nursing Support and Personal Care Workers (24.7%), and Ambulance Officers and Paramedics (22.6%).
The Human-Advantage Roles: A Role-by-Role Analysis
Nursing, Midwifery, and Allied Health
The nursing and midwifery workforce sits at the intersection of every structural AI-resistance factor: physical dexterity, emotional intelligence, clinical judgment, and regulatory accountability. These are not just safe jobs — they are urgently needed jobs.
The acute sector is predicted to experience the largest shortage, with projections indicating an undersupply of 26,665 FTE nurses by 2035, followed by the primary healthcare sector (21,765) and aged care (17,551). The broader picture is equally stark: the Australian Department of Health and Aged Care predicts a shortage of 85,000 nurses by 2025 and 123,000 nurses in Australia by 2030.
Midwifery specifically illustrates why emotional intelligence is not a "nice to have" — it is the core clinical function. A midwife assessing maternal distress, building birth plan trust, or managing a complication in a regional hospital without specialist backup is performing tasks that require embodied presence, real-time adaptive judgment, and therapeutic relationship — none of which AI can replicate.
Aged Care and Disability Support Workers
Australia's ageing population creates a structural demand floor that no amount of automation can eliminate. By 2031, nearly 20% of the population is expected to be aged over 65, up from around 16% now. The shortage of aged care workers will balloon to more than 400,000 by 2050 unless dramatic action is taken.
Australia will need at least 17,000 more direct aged-care workers each year in the next decade just to meet basic standards of care.
Despite raising wages up to 28.5% across the sector, Australia's workforce shortage in aged care will probably still reach 110,000 by 2030 according to some estimates.
Australia's healthcare system is experiencing significant strain as demand for aged care and disability services continues to grow. In 2026, the country faces mounting pressure from an ageing population, rising numbers of NDIS participants, and increasing regulatory expectations across care settings.
AI can assist with medication reminders, monitoring, and documentation. It cannot provide the physical care, human dignity, and relational presence that aged care residents require. Policy has reinforced this: the introduction of mandatory care minutes sets minimum targets for the amount of direct care residents must receive each day, including support delivered by registered nurses, enrolled nurses, and personal care workers.
Construction Trades and Labourers
The construction sector presents one of the most compelling cases for human-advantage employment through 2030. The numbers are extraordinary.
There is an estimated shortage of 141,000 construction workers, which could reach a peak of 300,000 by 2027. By occupational groups, shortages for Trades Workers and Labourers will peak at approximately 126,000, with the latter rising quickly.
Infrastructure Australia expects construction workforce demand to peak at 521,000 workers by 2027, well above last year's projection of 417,000 workers in 2026. The drivers are structural and long-term: much of this growth is driven by government-backed housing programs, energy transmission upgrades, and net-zero transition projects. Transmission projects alone account for $15 billion across the five-year outlook, while investment in social and affordable housing has climbed from $17 billion to $28 billion.
Australia faces a significant shortage of trades workers, especially in construction, driven by an ageing workforce, declining apprenticeship numbers, and surging demand from housing and infrastructure projects.
The automation argument for construction is particularly weak. Carpentry, plumbing, electrical work, and construction labouring all require fine motor dexterity, situational adaptation, and on-site problem-solving in environments that change daily. Carpentry, electrical, bricklaying, and plumbing — roles that demand manual dexterity and complex on-site decision-making — remain firmly human-centred.
Hospitality and Personal Services
The Victoria University simulation specifically flags hospitality as a growth sector in an AI-enabled economy. In a faster-growing economy, more people will work as teachers, hairdressers, and carers, because AI isn't expected to be as useful in those roles.
Hospitality is fundamentally about human experience. A skilled barista, a hotel concierge who reads a guest's mood, or a chef who adjusts a dish based on real-time feedback from the dining room are all performing tasks that combine physical skill, social intelligence, and creative judgment. While AI can optimise inventory and scheduling, it cannot replicate the hospitality experience that customers pay for.
The continued trend of growth in care and support occupations is expected to continue, with Community and Personal Services Workers expected to make up 11.4% of those employed in May 2033, compared to 11.1% in May 2023.
Mental Health and Social Work Professionals
The job growth outlook for Mental Health Professionals, including social workers, psychologists, and counsellors, indicates significant expansion. The broader category of Social and Welfare Professionals is expected to grow from 190,711 positions in 2023 to 218,967 by 2028, representing growth of approximately 28,256 jobs, or 14.8%.
The number of counsellor roles is projected to rise from 29,695 to 34,181 — an increase of about 4,485 jobs, or 15.1%. Similarly, social worker positions are forecasted to grow from approximately 43,311 in 2023 to 50,180 by 2028.
These roles involve navigating the most complex, ambiguous, and ethically fraught situations in human experience — precisely where AI's pattern-matching capabilities are least applicable and where human judgment, trust, and therapeutic presence are most essential.
Comparison: AI-Exposed vs. Human-Advantage Roles
| Occupation | AI Exposure (ILO) | JSA Growth Projection | Key Human Advantage |
|---|---|---|---|
| Data Entry Clerk | Very High (70% tasks automatable) | Negative/flat | None — high displacement risk |
| Nursing Support & Personal Care | Not Exposed | +24.7% to 2035 | Embodied care, clinical judgment |
| Dental Assistant | Not Exposed | +26.8% to 2035 | Physical dexterity, patient rapport |
| Construction Labourer | Low | +126,000 shortage by 2027 | Unstructured environments, physical skill |
| Social Worker | Low | +15.1% to 2028 | Ethical judgment, therapeutic relationship |
| Ambulance Officer/Paramedic | Not Exposed | +22.6% to 2035 | Emergency adaptability, physical intervention |
| Bookkeeper | High | Negative | Routine rules-based tasks — high displacement risk |
| Hospitality Worker | Low | Growing with AI-driven economy | Social intelligence, sensory experience |
Sources: ILO Exposure Indices (2024); Jobs and Skills Australia Employment Projections (2025); Infrastructure Australia Market Capacity Report (2025)
The CSIRO Insight: Why Domain Expertise Matters More, Not Less
CSIRO's Data61 — one of the world's largest applied AI research groups — has consistently found that AI does not eliminate the need for deep human expertise; it concentrates value in it. Knowledge in the form of human expertise is large, dynamic, and non-uniform — which is precisely why it is hard to codify and automate. A nurse who has spent a decade in oncology, a carpenter who has worked across heritage buildings, or a midwife specialising in high-risk pregnancies all possess contextual knowledge that cannot be extracted, formalised, and replicated by a language model.
This has a counterintuitive implication: as AI automates the routine components of professional work, the expert components become more valuable. AI may handle discharge documentation for a nurse, but it cannot replace the nurse's judgment about whether a patient is ready to go home. AI may generate a construction schedule, but it cannot replace the site foreman's read of whether the crew is safe to proceed in adverse conditions.
(For more on how AI augments rather than replaces expert workers, see our guide on AI and Australian Jobs Explained: Automation vs. Augmentation — What's the Real Difference?)
A Note on Nuance: Not All "Safe" Jobs Are Equally Secure
It is important to be precise. Not every job in a "safe" sector is equally protected, and not every worker in a growing occupation will automatically benefit.
JSA's employment projections are based on existing trends observed in the economy. They do not currently reflect the labour market implications of the adoption of generative AI and other emerging technologies. This means the projections likely understate the growth of human-advantage roles as AI drives economic expansion, while also underestimating disruption in some adjacent areas.
The equity dimension also matters. (See our guide on Who Is Most Vulnerable to AI Job Displacement in Australia? Gender, Age, Education, and Geography for the full analysis.) Entry-level positions within otherwise safe sectors — junior nursing assistants, entry-level hospitality workers — may face more pressure than senior, specialised roles. Geographic access to growing roles is also uneven, with regional workers facing additional barriers.
Key Takeaways
Care and construction dominate growth projections. Jobs and Skills Australia projects Nursing Support and Personal Care Workers to grow by 24.7% and Dental Assistants by 26.8% to 2035 — among the fastest-growing occupations in Australia.
Construction faces a structural labour crisis, not an automation threat. Infrastructure Australia's 2025 report projects a shortage of up to 300,000 construction workers by 2027, driven by a $242 billion public infrastructure pipeline. Trades workers and labourers face a peak shortage of 126,000.
The Victoria University 2050 simulation shows AI grows the economy — and with it, demand for human-intensive services. In an AI-enabled economy, more people work as carers, hospitality workers, and teachers precisely because AI cannot perform those roles.
Three structural barriers protect human-advantage roles: physical dexterity in unstructured environments, emotional intelligence and relational care, and contextual domain expertise applied to adaptive problems.
Aged care faces a shortage of 110,000 workers by 2030 despite significant wage increases — underscoring that demand for human care workers is driven by demographics, not just economics.
Conclusion
The AI-and-jobs debate is dominated by displacement narratives. The evidence — from Victoria University's 2050 simulation, Jobs and Skills Australia's occupational projections, and Infrastructure Australia's workforce capacity reports — tells a more complex and ultimately more optimistic story. AI is not simply taking jobs. It is reshaping which jobs the economy needs, and the roles that require physical presence, emotional intelligence, and embodied expertise are growing faster than ever.
For Australian workers, this means the most durable career investments are not necessarily the most technologically sophisticated ones. A registered nurse, a licensed electrician, a midwife, or a skilled construction worker is entering a labour market that will be chronically short of their skills for the next decade, regardless of what AI does to office-based roles.
The challenge is not that these jobs are disappearing. The challenge is that Australia is not producing enough people to fill them.
For workers considering their next move in an AI-disrupted market, see our guides on Should You Retrain, Pivot, or Stay? How to Decide Your Best Career Move in an AI-Disrupted Australian Job Market and How to Future-Proof Your Career Against AI in Australia: A Step-by-Step Upskilling Plan.
References
Jobs and Skills Australia. "Employment Projections — Outlook by Occupation." Jobs and Skills Australia, 2025. https://www.jobsandskills.gov.au/data/employment-projections/occupation
Jobs and Skills Australia. "2025 Employment Projections Summary." Jobs and Skills Australia, December 2025. https://www.jobsandskills.gov.au/sites/default/files/2025-12/employment_projection_-_summary.pdf
Jobs and Skills Australia. "Future Workforce Needs and Growing Sectors in Australia." Jobs and Skills Australia, 2025. https://www.jobsandskills.gov.au/news/future-workforce-needs-and-growing-sectors-australia
Dixon, Janine and Lennox, James. "These Jobs Will Thrive — But Others May Vanish — as AI Transforms Australia's Workforce." The Conversation / Victoria University, August 2025. https://www.vu.edu.au/about-vu/news-events/news/these-jobs-will-thrive-but-others-may-vanish-as-ai-transforms-australias-workforce
Infrastructure Australia. "2025 Infrastructure Market Capacity Report." Infrastructure Australia, November 2025. https://www.infrastructureaustralia.gov.au/reports/2025-infrastructure-market-capacity-report
Australian Primary Health Care Nurses Association (APNA). "Aged Care Nurses Underutilised Despite Workforce Shortage — Survey." APNA, 2022. https://www.apna.asn.au/about/media/archive-media-releases/aged-care-nurses-underutilised-despite-workforce-shortage---survey
Committee for Economic Development of Australia (CEDA). "Australia's Dire Shortage of Aged-Care Workers Requires Immediate Action." CEDA, 2021. https://www.ceda.com.au/news-and-resources/media-releases/health-ageing/australia%E2%80%99s-dire-shortage-of-aged-care-workers-req
Lowy Institute. "Amid a Global Aged Care Labour Shortage, How Will Australia Address the Challenge?" Lowy Institute, 2024. https://www.lowyinstitute.org/the-interpreter/amid-global-aged-care-labour-shortage-how-will-australia-address-challenge
CSIRO Data61. "Artificial Intelligence Roadmap." CSIRO, 2019. https://www.csiro.au/-/media/D61/Reports/AI-Roadmap/19-00346_DATA61_REPORT_AI-Roadmap-7.pdf
Australian Nursing and Midwifery Journal (ANMJ). "Australia Facing Shortfall of Over 70,000 Nurses by 2035, Report Reveals." ANMJ, August 2024. https://anmj.org.au/australia-facing-shortfall-of-over-70000-nurses-by-2035-report-reveals/
Jobs and Skills Australia. "Employment Projections for the Decade Ahead." Towards a National Jobs and Skills Roadmap, 2023. https://www.jobsandskills.gov.au/publications/towards-national-jobs-and-skills-roadmap-summary/employment-projections-for-the-decade-ahead