Australia's central bank finds AI hitting young workers first, especially in software and customer service

Australia's central bank has published an analysis by Mike Major that sets up a monitoring system for where AI's labour-market effects should surface first: entry-level workers in occupations with high potential exposure, such as programming, clerical work and customer service. The headline finding is that job postings have declined across every group since ChatGPT launched in late 2022, but the fall is steepest in the highest-exposure quintile. Within that group, workers aged 15 to 24 and recent college graduates show weaker employment rates and higher unemployment than older workers. In high-exposure jobs, youth employment has drifted down since late 2022 while employment among older workers has risen.
Major is explicit that this does not establish causation. The exposure measure captures theoretical potential for AI to affect a job, not whether companies in those roles actually use AI. Post-pandemic shifts in the business cycle and changes in remote work patterns are named as plausible alternative explanations for the numbers. Data from the job marketplace SEEK is cited as showing AI complementing existing work while also generating new demand for AI skills, which cuts against a simple displacement story.
For developers, the detail worth holding onto is the split between age bands inside the same occupation. A junior software developer and a senior engineer sit in the same exposure quintile, yet the Australian pattern puts them on opposite trajectories. This suggests the adjustment, if AI is driving it, is happening at the hiring margin rather than through layoffs of existing staff. That reading is my own interpretation of the reported figures, not a claim the analysis makes.
Other research points the same way, with its own caveats. A Stanford study from August 2025, drawing on millions of US payroll records, found employment among 22- to 25-year-olds in AI-exposed occupations down 13 percent in relative terms since late 2022, and nearly 20 percent among young software developers, while more experienced workers in the same occupations held steady or gained. A March 2026 Anthropic study also reported weaker hiring for 22- to 25-year-olds in exposed occupations, with job-finding rates about 14 percent lower, but found no systematic rise in overall unemployment. Its conclusion was that theoretical automation potential and actual AI use remain far apart.
Anthropic's own company is offered as a concrete example. An internal review found the share of junior hires in technical roles fell from 27 percent to 9 percent since early 2025, and in August 2026 AI agents across the company were logging six times as many hours as human employees. Anthropic acknowledges it is a special case as an unusually AI-intensive software company. Researchers nonetheless warn of a "tragedy of the cognitive commons": if junior roles disappear, the pipeline that produces experienced professionals thins out, assuming those professionals are still needed. For engineering organisations, that is a succession problem rather than a headcount problem, and it plays out over years.
The economy-wide picture is quieter. A study of 25,000 workers in Denmark found a 64 percent AI adoption rate but no measurable change in wages or hours worked, with workers saving an average of just 2.8 percent of their working time. A January 2026 analysis of 10.6 million LinkedIn profiles and US unemployment data, led by Morgan Frank at the University of Pittsburgh, complicates the causal story further: trouble in exposed occupations began in early 2022, months before ChatGPT arrived, pointing to macroeconomic factors such as the Fed's interest rate shift or a correction after pandemic-era hiring surges.
Demand is also being created in parallel. The World Economic Forum's Future of Jobs Report 2025 found 77 percent of surveyed companies prioritising upskilling over job cuts and projects a net gain of 78 million jobs by 2030, while LinkedIn reports AI-related job postings grew sixfold in a year. These figures are projections and platform observations rather than measured outcomes, and they sit awkwardly beside the entry-level declines, which is why the RBA frames its work as monitoring rather than a verdict.
Why it matters: teams that hire juniors into software, support and administrative roles should treat this as a signal to watch their own entry-level pipelines rather than a reason to change course, since the RBA data is correlational and the strongest single-company evidence comes from an acknowledged outlier. The practical risk, in my reading, is a quiet loss of the apprenticeship layer that turns graduates into senior staff. Product and engineering leaders may want to track junior hiring and ramp-up time as metrics in their own right, and pair agent tooling with deliberate mentoring rather than assuming experience will accumulate on its own.
Based on reporting from the original publisher. Visit the source for full context and later updates.
Publisher excerpt
Australia's central bank sees early signs that AI is hitting young entry-level workers in highly exposed jobs like software development and customer service. Since ChatGPT launched, job postings and employment among 15- to 24-year-olds in those fields have dropped more sharply than among older workers. Author Mike Major cautions against jumping to conclusions, however, noting that business cycle shifts and remote work trends could also explain the numbers. The article Australia's central bank finds AI hitting young