AI exposure and the labour market
Read this first. Exposure is not displacement. A high exposure score means Gen AI could perform or assist many of an occupation's tasks. It does not mean jobs are being lost. This page describes trends side by side and makes no causal claim. Four concepts are kept separate throughout: theoretical task exposure, observed AI usage, employer demand for AI skills, and employment and job-ad outcomes.
Job ads, indexed to a baseline
Online job ads, index (baseline = 100)
Absolute counts are shown alongside, on separate axes, because the two series differ in scale by a factor of several hundred.
Three-month baselines use the published observation ending in the selected month. The 12-month baseline averages four non-overlapping three-month observations. Single-month counts cannot be recovered from these data. If a required baseline observation is missing, that series' index is unavailable.
The IVI 4-digit data is a 3-month moving average of original, not seasonally adjusted, counts. Unit groups are not summed: the 3-month averages of the occupations do not add up to the published total. ANZSCO 5996 is a unit group, and this is the finest level the IVI measures. Its 6-digit roles are not separately measured.
Gen AI exposure of insurance-relevant occupations
Source: JSA Gen AI Capacity Study · Score version:
Coming in milestone 3: the full insurance watchlist, with job ads, employment, exposure, measurement granularity and mapping coverage for each role, plus the exposure-versus-change scatter. Coming in milestone 4: the AI-mention share of job postings (Indeed Hiring Lab) and alternative exposure indices as sensitivity checks.