Key points
- An RBA analysis by Mike Major finds job postings fell in every occupation group since late 2022, with the sharpest drop in the most AI-exposed roles.
- Within those roles, people aged 15 to 24 and recent graduates show weaker employment and higher unemployment than older workers.
- The study does not prove causation and measures theoretical exposure, not how much companies actually use AI.
- A University of Pittsburgh analysis found trouble in exposed occupations began in early 2022, before ChatGPT launched.
Why it matters for marketers: Entry-level hiring in writing, support and software-adjacent roles may be shrinking first, so teams should plan for thinner junior pipelines while avoiding assumptions that AI alone is the cause.
Researchers at Australia's central bank have built a monitoring approach aimed at spotting where AI's labor effects would appear first. The analysis, written by Mike Major, looks at junior workers and at occupations seen as highly exposed to AI, including programming, clerical work and customer service. Overall, AI has displaced very few jobs in Australia so far.
Still, the early signals point to young people. Postings have declined across all groups since late 2022, and the largest decline is in the most exposed category. Within it, the youngest workers and recent graduates fare worse than older colleagues. Major cautions that the work cannot show cause and effect, and that business cycle changes or remote work trends could play a part. SEEK data suggests AI is also complementing existing work and generating demand for AI skills.
Other research is mixed. A Stanford study and an Anthropic study both found weaker hiring of early-career workers in exposed fields, while a Danish study of 25,000 workers found no economy-wide change in wages or hours. The Pittsburgh team led by Morgan Frank found that the slump started months before ChatGPT arrived, which points to macroeconomic forces.
For marketing and content leaders, the takeaway is about talent planning. Junior roles in copy, support and technical production may be the first to thin out, and that could leave a shortage of experienced people later. Watch for data on actual AI use inside companies, which would show whether exposure is turning into real substitution.
Sources
This brief is our own summary. Read the full reporting at the source.