Will AI Take Your Job? What Work Could Look Like in 2031 and 2036

Published 25 October 2026 | NEXTNEWS Future of Work Special Report. Will artificial intelligence replace your job over the next five to ten years? For many people, the more useful question is which parts of the job will change, who will benefit, who could lose bargaining power and what skills will remain valuable. The answers are not uniform: an administrative assistant, speech pathologist, accountant, electrician, software developer and delivery driver all face different kinds of automation, risk and opportunity.

By 2031, many jobs could incorporate AI-based assistants as routinely as spreadsheets or email today. By 2036, advanced systems could organise and execute more complex workflows, subject to human supervision, legal safeguards, cost and technological limits. These are plausible scenarios—not predictions of specific inventions or guaranteed job losses.

Public evidence suggests both disruption and opportunity. The World Economic Forum’s widely cited 2025 employer-based outlook anticipated 170 million jobs created and 92 million displaced globally by 2030, a net increase of 78 million. Crucially, the estimate covers multiple drivers, including demographics, economic change and the green transition, not just AI. Australia’s own July 2026 government research found no evidence of broad AI-driven employment upheaval yet, while detecting signs of slower growth in some occupations exposed to generative-AI automation. WEF · Australian Department of Employment and Workplace Relations.

Key findings: AI’s impact on jobs in five to ten years

• Tasks will generally change before job titles disappear. A role combines repeatable tasks, human judgement, responsibility, relationships, physical activity and professional regulation.
• Routine digital work is more exposed. Standardised writing, data extraction, customer enquiry triage, transcription, document review and some software tasks can increasingly be automated or accelerated.
• Exposure does not equal unemployment. Employers can use saved time to serve more customers, improve quality, reduce hours, change staffing, lower costs or expand services; the outcome depends on demand and business decisions.
• Care, trades, education and complex field work will change rather than simply disappear. AI can help with paperwork or decision support, but trust, safeguarding, physical work, empathy and accountability remain important.
• Entry-level pathways are a concern. If AI takes over basic drafting, coding and analysis, employers will need alternative ways for graduates and apprentices to gain experience.
• Workers who can verify and safely apply AI may gain. Digital literacy is becoming a baseline, while sector expertise, communication and judgement help differentiate people.

What the best research actually says — and what it does not

Source documents: WEF; IMF; ILO; JSA; DEWR. Different research approaches, baselines and definitions mean these figures should not be added together or interpreted as directly comparable probabilities for a particular job.

What happens by 2031? A plausible five-year workplace

In a realistic 2031 workplace, many organisations may operate with AI embedded inside existing software: email drafting, meeting summaries, document search, invoice reconciliation, customer-service routing, basic translation, fraud alerts, early clinical decision support, scheduling and software coding. Some staff could manage several AI tools or workflow agents instead of entering each item manually.

Imagine a small allied-health clinic: an administrator may spend less time sending appointment reminders and compiling reports, but more time checking consent, resolving exceptional billing claims, communicating with families and arranging complex bookings. A speech pathologist or occupational therapist might use carefully approved AI tools to organise non-identifying notes or suggest education resources while retaining responsibility for diagnosis, clinical decisions, privacy and the therapeutic relationship. That example is an illustrative scenario, not evidence that a particular clinic has achieved it.

For a mortgage brokerage, software might collect documents, flag missing information and prepare comparisons, but licensed professionals remain responsible for assessing suitability, explaining risks and complying with credit obligations. A skilled electrician could use AI for estimates, fault-finding prompts and scheduling, while still performing licensed on-site work.

In each case, the business model may change even when the profession survives. Organisations could hire fewer people for narrowly defined repetitive duties, while creating demand for workers who manage exceptions, analyse outputs, protect clients and lead redesign. Some junior roles may become harder to enter unless employers redesign training.

What about 2036? Three different 10-year scenarios

No credible researcher can identify the exact employment distribution in 2036. Consider three conditional futures rather than one confident prediction:

Scenario A: Productivity with broad-based job redesign

AI becomes reliable and affordable for many repetitive digital workflows, but demand for health care, education, housing services and new products expands. Firms redeploy employees, shorten turnaround times and invest in training. Some clerical positions shrink, yet employment adjusts through new specialties and expanded services. This is a potential high-adoption, high-complementarity pathway.

Scenario B: Uneven transition and a squeezed career ladder

Leading companies use advanced systems to deliver more output with leaner junior teams. Hiring slows for some administrative, analytical and early-career digital positions. Experienced staff who understand the business and can supervise AI become more productive; graduates struggle to acquire the experience that used to come from routine assignments. Wage and opportunity gaps widen without deliberate apprenticeship and training reforms.

Scenario C: Slower adoption than the technology headlines suggest

Progress continues but companies face privacy restrictions, cyber risk, poor source data, inaccurate model outputs, cost, energy demand, worker resistance and liability concerns. Highly regulated activities adopt AI unevenly, and smaller businesses cannot reliably capture all the promised savings. Job changes occur gradually, with traditional human-led workflows still common.

The actual labour market could combine these scenarios across sectors. Changes in immigration, population ageing, productivity, the business cycle, public policy and trade will affect hiring alongside AI.

Jobs most exposed to change: a sector-by-sector outlook

The table ranks broad tasks and possible redesign pressures—not individual probabilities of being made redundant. It is NEXTNEWS’s qualitative synthesis of the sources above, not a statistical forecast.

*Qualitative editorial scenarios, not an individual risk score, occupation-specific forecast or Australian government classification. In practice a firm’s data, customer demand, organisational design, regulation and worker bargaining power determine what happens.

Which jobs could grow as AI spreads?

Directly technology-related roles could include AI systems integration, machine-learning operations, data engineering, evaluation and testing, cybersecurity, model-risk assurance, responsible-AI governance and digital accessibility. But focusing only on “prompt engineer” job titles is too narrow. Many organisations will need experienced industry specialists who can identify safe applications, validate results and redesign services.

Some roles may expand because of structural forces alongside AI: aged care, disability support, nursing and therapy, teaching, renewable-energy infrastructure, construction, electrical work, logistics and maintenance. The WEF report highlights both advanced technical roles and essential community-economy occupations as sources of future labour demand. Its overall job projections should not be mistaken for proof that AI creates every new role.

There could also be new hybrid careers: an occupational therapist proficient in accessible AI tools, a bookkeeper who audits automated ledgers, an administrator who supervises workflow agents, or a teacher who creates inclusive personalised learning under human oversight. Job titles may change less than the mix of tasks inside them.

What could happen to Australian salaries and recruitment?

Demand is already changing, but salary promises require care. PwC Australia’s June 2026 AI Jobs Barometer reported that Australian job advertisements requesting AI skills rose from approximately 20,000 in 2024 to 41,000 in 2025, and it estimated an average 62% wage premium associated with AI-skilled roles in its dataset. This is not a guarantee that learning an AI tool causes an individual salary to rise 62%. Occupation, seniority, location, education, employer and demand can all affect the observed premium.

At the other end of the market, straightforward digital-production tasks may face more price competition as software makes output easier and as businesses globalise sourcing. The crucial divide may be less “technology worker versus non-technology worker” than people able to combine knowledge with tools versus roles offering undifferentiated routine output.

For Australian employers, existing worker obligations continue to apply. Decisions about surveillance, workplace privacy, discrimination, awards, enterprise agreements, occupational health and safety and consultation cannot simply be delegated to an opaque algorithm. Related NEXTNEWS analysis: AI and workplace surveillance.

The entry-level job problem: how does the next generation learn?

One risk gets less attention than headline layoffs: the first rung on the career ladder. In law, accounting, software, design and consulting, repetitive junior assignments have historically helped people learn how the work is done. If AI performs those assignments cheaply, companies may eliminate the entry-level position without replacing the educational value it provided.

Employers can respond by designing structured apprenticeships and supervised learning, including review of incorrect AI results, client conversations, ethical decisions, practical assignments and rotations across departments. Graduates can document credible applied work, explain the limits of tools they use and demonstrate that they can verify results rather than merely produce polished AI-generated text.

Australia’s Jobs and Skills Report 2026, released 12 October, says approximately 29% of occupations were experiencing skills shortages in 2025. Labour shortages in some sectors can coexist with hiring pressure elsewhere; an economy can have vacancies while particular graduates struggle to enter their chosen professions.

Will AI benefit older workers or make it harder to stay employed?

Both outcomes are plausible. Tools that simplify writing, make specialist knowledge easier to retrieve, reduce repetitive paperwork or improve accessibility could help mature workers continue contributing. But unplanned software rollouts can exclude people with limited training, disabilities, unreliable access or highly specialised older workflows.

Employers should provide role-specific training and accessible alternatives rather than assuming workers can master new systems independently. Workers who have spent decades building judgement, networks or practical expertise may hold advantages that complement the speed of AI tools.

AI in Australia versus the rest of the world

Australia has distinctive labour-market conditions: a large service economy, growing health and care needs, regulated professional services, major construction and resources sectors, and geographic distance between communities. Adoption also depends on reliable digital infrastructure and the cost of deploying technology across smaller businesses.

Internationally, the IMF estimated about 60% of employment in advanced economies is exposed to AI, compared with roughly 40% in emerging markets and 26% in low-income countries. Those are exposure estimates, not job-loss predictions. Lower exposure does not mean less risk for workers dependent on outsourced digital services; nor does higher exposure guarantee that advanced economies lose more jobs. Differences in skills, labour costs, infrastructure and sector composition shape outcomes.

For business-process outsourcing, cross-border administration and digital freelancers, AI can compress routine work but also create demand for multilingual service quality, secure data handling, integration, compliance, advanced client support and outcomes-based work. A purely hourly data-entry business may face pressure, while a specialist outsourced team that validates AI-assisted claims or supports regulated processes may be able to differentiate itself. Those are business scenarios, not earnings forecasts.

What should workers learn now? A practical 12-month plan

1. Map your job into tasks. Identify which activities are repetitive and digital, which require professional judgement, and which rely on relationships, physical presence or legal responsibility.
2. Choose an appropriate AI tool and work within your employer’s rules. Start with low-risk information, not confidential medical, payroll, financial or customer records. Never enter protected data into unapproved systems.
3. Learn verification. Test whether generated summaries accurately reflect sources, calculations agree with records, code passes tests and outputs contain bias or fabricated claims.
4. Build sector expertise. Understand the actual workflow: NDIS compliance, allied-health administration, accounting controls, cybersecurity, construction safety, care planning, legal duties or the field relevant to you.
5. Strengthen transferable skills. Focus on problem definition, writing clearly, negotiating, coaching, communication, data analysis and sound decision making.
6. Create a portfolio of work. Document measurable improvements you delivered, including your checking process and what remained under human supervision. Never claim automated results you did not verify.
7. Review your career path every six months. Track changes to job advertisements, required skills, relevant certifications and mentoring opportunities, rather than following one prediction about “jobs AI cannot replace”.

What employers and policymakers can do to avoid a damaging transition

Businesses can audit tasks before headcount decisions, evaluate error and security costs, consult workers, train supervisors, clarify accountability and measure customer results rather than merely counting automated actions. Efficiency gains can be invested in new services, better wages, reduced burnout or workforce development; they need not translate automatically into redundancies.

Governments, education institutions and industry bodies can improve affordable mid-career training, digital accessibility, apprenticeship pathways, career-transition services, workforce data and transparency around automated decisions. Without safeguards, the economic benefits may be concentrated in firms that own systems and data, leaving workers with weaker bargaining power.

AI systems can also fail. Incorrect advice, hallucinated references, cybersecurity vulnerabilities, biased recruitment filters, privacy breaches and unreliable integration can destroy value. Any decision affecting healthcare, lending, employment or rights requires controls proportionate to the risk and compliance with applicable Australian rules.

Questions to ask before choosing a career for 2031 or 2036

Rather than asking “Which profession is guaranteed safe?”, consider four questions: What real problem does this work solve? How much of the work can be fully specified and checked digitally? Who is legally and socially responsible when something goes wrong? And will demand for the service expand if its cost falls?

Many future jobs will combine repeatable processes with complex human responsibilities. A career may be vulnerable in its entry-level tasks yet attractive for its more advanced specialities. Equally, a highly technical title could change quickly as tools improve. The sensible response is to develop adaptable expertise, not to assume any single qualification guarantees lifetime security.

FAQ: AI and the future of employment

Will AI replace millions of jobs by 2030?

Many roles are likely to change and some will be displaced, but there is no single agreed AI-specific displacement number. The WEF projects 92 million jobs displaced and 170 million created from all major structural labour-market trends by 2030. It does not say AI alone destroys 92 million jobs.

Which jobs are most at risk of AI automation?

Work composed largely of predictable, repeatable digital tasks may experience the greatest pressure. That includes some clerical processing, simple content production and routine customer-service activity. Exposure does not mean every worker in those fields will lose their job.

What jobs are least likely to disappear?

Roles requiring complex physical tasks, interpersonal trust, care, real-world troubleshooting, safeguarding or accountable judgement are less straightforward to automate end-to-end. However, AI may still transform their administration, diagnosis support, planning and communication. No career is completely “AI-proof”.

Will AI increase wages or reduce them?

It may do both in different sectors. Scarce high-value skills may command higher pay, while automatable commodity tasks may face greater competition. Organisational choices, wages law, demand, bargaining power, education and economic conditions will affect outcomes.

Should students still study programming, law or accounting?

These fields still have important human responsibilities, but their teaching and junior work may evolve rapidly. Students should combine fundamentals with practical AI literacy, ethics, verification and work placements rather than rely solely on routine techniques.

Is 2036 employment predictable today?

No. Changes in AI reliability, business adoption, regulation, demographics and the broader economy make point forecasts inherently uncertain. The 2031 and 2036 examples in this feature are explicitly scenario analysis.

Sources, methodology and editorial transparency

• World Economic Forum: Future of Jobs Report 2025 (global structural projections to 2030)
• IMF: AI and global labour-market exposure, 2024
• ILO–NASK: global GenAI exposure index, May 2025
• Jobs and Skills Australia: GenAI Capacity Study, 2025
• Department of Employment and Workplace Relations: AI and employment in Australia, July 2026
• Jobs and Skills Australia: Jobs and Skills Report 2026, October 2026
• PwC Australia: 2026 AI Jobs Barometer (job-ad and wage-association analysis)

How we prepared this article: This report distinguishes published observations and attributed projections from NEXTNEWS’s own illustrative scenario analysis for 2031–2036. The role-sector assessment is editorial synthesis of task characteristics and the cited research; it is not an employer survey, individual redundancy assessment, proprietary econometric forecast or job-security rating. The two in-article images and the featured image were generated using AI as conceptual editorial illustrations and are labelled accordingly. They portray no actual company, employees or events.

Disclaimer: NEXTNEWS provides general news, commentary and educational information, not individual employment, career, legal, financial, investment, education, migration or business advice. AI and labour-market forecasts are uncertain and may change quickly. No job can be guaranteed safe or certain to disappear, and no income, salary premium, training outcome or redundancy prediction is promised. Before making employment, study, financial or business decisions, assess your personal circumstances and seek appropriately qualified professional advice. Australian legal and workplace requirements vary with industry, location and employment arrangement. Information verified as at 25 October 2026.