The debate about robots and jobs has been running for a decade, and it has produced more heat than light. The "robots will take all the jobs" camp has consistently overestimated the pace of automation. The "automation creates more jobs than it destroys" camp has consistently underestimated the disruption to specific workers and communities. The reality, as the data from 2026 is beginning to show, is more nuanced than either narrative.

What the Data Actually Shows

The McKinsey Global Institute's 2026 automation report — the most comprehensive analysis of automation's labour market impact — finds that 30% of tasks across the global economy are now technically automatable with current technology. But "technically automatable" and "economically automated" are very different things. The actual automation rate — the share of tasks that have been automated in practice — is 12%, up from 7% in 2020.

The gap between technical possibility and economic reality reflects the genuine complexity of automation deployment: the cost of robots and AI systems, the difficulty of integrating them into existing workflows, the regulatory environment, and the social and political resistance to workforce displacement. Automation is happening — but more slowly, and more unevenly, than the most alarming predictions suggested.

Which Jobs Are Actually at Risk

The jobs most vulnerable to automation share common characteristics: they involve repetitive, predictable tasks; they can be performed in controlled environments; and they don't require social intelligence, creative problem-solving, or physical dexterity in unstructured settings. The most at-risk occupations in Australia include: data entry clerks (85% automation probability), telemarketers (83%), bookkeeping clerks (78%), and assembly line workers (72%).

The jobs least at risk are those that require human connection, creative judgement, or physical adaptability in unpredictable environments: mental health professionals (0.3% automation probability), primary school teachers (0.4%), social workers (0.3%), and skilled tradespeople like electricians and plumbers (2–4%).

The Augmentation Story

The more accurate framing for most workers isn't replacement but augmentation. AI and robotics are changing the nature of work rather than eliminating it. Radiologists now work alongside AI systems that flag anomalies for human review — the AI handles the routine screening, the radiologist handles the complex cases. Lawyers use AI for document review and legal research, freeing them for client work and courtroom advocacy. Construction workers use exoskeletons that reduce physical strain and improve precision.

In each case, the human worker is more productive with the technology than without it — and the technology is more valuable with human oversight than without it. This augmentation model is the dominant pattern in the current wave of automation, and it suggests a more optimistic medium-term outlook than the replacement narrative implies.

The Transition Challenge

The optimistic aggregate story conceals a difficult distributional reality. The workers most at risk from automation are disproportionately older, less educated, and concentrated in specific geographic areas and industries. The new jobs created by automation — robot technicians, AI trainers, data analysts — require different skills and are often in different locations. The transition from displaced worker to new-economy worker is not automatic, and it requires investment in retraining, income support, and geographic mobility that most countries are not yet providing at adequate scale.

Policy Responses: What Works

The countries that have managed automation transitions most successfully share common policy features: strong social safety nets that provide income support during transitions, active labour market policies that fund retraining and job placement, and education systems that emphasise adaptability and lifelong learning over narrow vocational training.

Denmark's "flexicurity" model — which combines easy hiring and firing with generous unemployment benefits and active retraining — is frequently cited as a model for managing automation transitions. Germany's apprenticeship system, which creates strong pathways from education into skilled trades that are difficult to automate, has helped maintain low unemployment despite significant manufacturing automation. Australia's approach — relatively weak social safety nets, limited retraining investment, and a skills system that has struggled to adapt to changing labour market needs — is less well-positioned for the transition ahead.

The 10-Year Outlook

The most credible forecasts suggest that automation will displace 15–20% of current jobs in Australia over the next decade — but create a similar number of new jobs in the process. The net employment effect will be roughly neutral, but the transition will be painful for workers in the most affected occupations and regions. The policy challenge is not preventing automation — which would sacrifice the productivity gains that fund higher living standards — but managing the transition in a way that distributes the benefits broadly and supports those who bear the costs.

Sources & Further Reading