IMF puts Australia at the top of the rise in unemployment in the developed world, and AI is the factor no one knows how to measure

By 1News · October 18, 2026.

Australia is on track to post the biggest rise in unemployment in the developed world, according to new projections from the International Monetary Fund (IMF). The rate would climb from an average of 4.4% this year to 4.7% in 2027, with some 100,000 jobs at risk, New Zealand broadcaster 1News reports.

The contrast is stark. The IMF says most economies have weathered the turbulence of the past six months, and Australia is the exception. New Zealand, with higher unemployment (5.4%), expects it to fall to 5.1% next year. According to 1News, the new figures also suggest the talent drain to Australia is slowing: New Zealand's net migration loss was 27,900 people in the year to March, down from more than 31,000 the previous year (provisional data).

IMF deputy director of research Petya Koeva Brooks said the global economy has been "very resilient" but may be reaching "the limits of that resilience". The IMF expects artificial intelligence to be the main disruptive force, and Brooks added that "a fairly small number of countries" are benefiting from the AI-driven boom.

An important caveat: the article does not say AI explains the deterioration in Australian employment. It links the two topics but presents no figures connecting them. One account, from a New Zealander living in Australia who describes a market with many temporary positions and few full-time ones, points more to a problem of job quality than to any specific technological displacement.

Our take: what matters is not Australia but the thesis the IMF hints at. If the AI boom is concentrated in a few countries, the rest may go through the transition without collecting the dividend: they absorb the labor uncertainty without hosting the data centers, the labs or the capital. It is a form of inequality between nations that adds to the inequality between sectors.

In that context, the policy response cited in the article is telling. A joint parliamentary committee on AI is calling for immediate investment and looser copyright laws so Australian companies can train models on local and international content. It is a bid not to fall behind, but it raises a fundamental debate about who pays for the training: the creators of that content.

In the short term, simple explanations deserve suspicion. A three-tenths-of-a-point rise in unemployment is modest and may come down to interest rates, consumer spending or the business cycle rather than automation. In the long term, our thesis holds: the transition will be uneven and hard for those who go through it without a safety net, but the abundance AI promises will reach everyone only if the countries left out of the boom today manage to take part in it. For that, talent, infrastructure and distribution policy matter more than any unemployment projection.