Report: AI struggles to replace nurses, teachers and labourers
AI struggles to replace nurses, teachers and labourers; a 2026 report finds hands-on roles cost several times more to automate than paying human wages.
Not all work can be coded away. A new analysis indicates that while algorithms are improving fast, replacing people in care, classrooms and construction is still prohibitively expensive.
A June 2026 report by Planera compared automation costs with median wages and found that many frontline roles would cost employers several times more to automate than to staff with humans.
Why replacing frontline workers is still too expensive
The report found some of the largest cost gaps in professions that rely on physical work and human interaction. Automating the role of a nursing assistant, for example, is estimated to cost around $375,100 a year, compared with a median annual salary of about $42,200.
Home health and personal care aides would cost roughly $282,200 to automate, while construction labourers would require automation costing around $285,300 a year. Teaching assistants also remain difficult to replace, with estimated expenses of nearly $194,300 compared with average annual wages of about $36,700.
These figures suggest that, at current technology and hardware costs, employers have little financial incentive to replace workers in these occupations.
Routine tasks remain easier to automate
The picture looks very different for jobs involving repetitive and predictable tasks. The report identifies cashiers as one of the most economically viable roles for automation. Self-checkout systems are estimated to cost around $24,220 annually, lower than the typical cashier's median wage of about $32,880.
In short, tasks that follow fixed processes and require limited decision-making are generally easier and cheaper to automate than jobs involving unpredictable environments or regular interaction with people.
Why AI still struggles in physical jobs
Many frontline roles require skills that AI and robots have yet to master. Caring for patients, navigating busy workspaces or handling irregular objects demands physical dexterity, situational awareness and sound judgement.
Beyond the technology itself, employers also face costs related to installation, maintenance, safety compliance, insurance and employee training. These additional expenses often outweigh the potential savings from automation.
Digital jobs face growing pressure
Interestingly, the report suggests that some white-collar professions may be more exposed than physical jobs. As AI coding assistants continue to improve, software development is becoming increasingly susceptible to automation, with some large technology companies already reducing engineering roles.
For now, the findings suggest that AI is more likely to complement workers than replace them outright. While automation will continue expanding, jobs that depend on adaptability, practical problem-solving and human interaction remain among the hardest and most expensive to automate.