Augmented, not replaced?

The real map of generative AI and Australian jobs

Published

June 12, 2026

Ask an Australian worker what generative AI means for their job and you tend to hear one of two answers: nothing much, or everything. Mapped across the whole workforce, the truth is more specific - and more hopeful - than either.

When Jobs and Skills Australia, the federal government’s labour-market agency, modelled generative AI against every occupation in the country, it measured two very different things. How much of a job AI could automate - do instead of you. And how much it could augment - help you do. Those are not the same number, and the gap between them is the story.

For 357 occupations covering 14.4 million Australians, this is how much of the average job generative AI could take on.

The red curve is automation - tasks AI could do alone. For most workers it tops out around 34% of their job.

The dark curve is augmentation - tasks AI could help with. It sits far to the right: a median of 66%. AI is a far better assistant than it is a replacement.

That gap reframes the whole anxious debate. So let’s put every occupation on the same map - automation across the bottom, augmentation up the side - and see where Australian work actually lands.

Every bubble is one occupation, sized by how many Australians do it. The dashed line is the tipping point: below it, automation wins; above it, augmentation wins.

Almost the entire workforce sits above the line. On an employment-weighted basis, 97% of Australian workers are in jobs AI is more likely to assist than to replace.

Only 8 of 357 occupations fall below the line into genuine automation territory. The robots-take-all headline describes a sliver of the labour market.

Colour marks the share of women in each role. Hover any bubble - the redder ones, women’s work, do not sit safely in the corner.

Augmentation may be the rule, but exposure is not shared equally. To see who carries the automation risk - the kind that substitutes for a worker rather than helping one - line every occupation up by the share of women who do it.

Each bubble is an occupation again, now placed by how female it is (across) and how automatable (up). The trend tilts upward: the more women in a job, the more of it AI could do alone.

The red cluster is the sharp end - 288,000 General Clerks (83% women, 71% automatable), 176,000 Receptionists (92% women), 132,000 Accounting Clerks (78% women).

Women’s jobs average 40% automation exposure against 35% for men - even though their augmentation upside is almost identical. The International Labour Organization finds the same skew worldwide.

So the question generative AI poses to Australia is not whether the machines are coming for our jobs. For all but a handful of occupations, they are coming to help. The real question is whether the workers in the most-exposed roles - the clerks, the receptionists, the women who hold them - get the training, the tools and the time to do the augmenting themselves, before someone decides the cheaper path is to automate instead.

That is a policy choice, not a foregone conclusion. The data says Australian work can be augmented. Whether it is - that part is still up to us.

Data sources & references

  • Anthropic. (2026). Claude (Opus 4.8) [Large language model]. https://claude.ai
  • Australian Bureau of Statistics. (2026). Labour force, Australia, detailed (cat. no. 6291.0.55.001, Table EQ08: Employed persons by occupation unit group of main job, sex, state and territory) [Data set]. https://www.abs.gov.au/statistics/labour/employment-and-unemployment/labour-force-australia-detailed/mar-2026
  • Gmyrek, P., Berg, J., Kamiński, K., Konopczyński, F., Ładna, A., Nafradi, B., Rosłaniec, K., & Troszyński, M. (2025). Generative AI and jobs: A refined global index of occupational exposure (ILO Working Paper No. 140). International Labour Organization. https://www.ilo.org/publications/generative-ai-and-jobs-refined-global-index-occupational-exposure
  • Jobs and Skills Australia. (2025). Our gen AI transition: Implications for work and skills [Data set]. Commonwealth of Australia. https://www.jobsandskills.gov.au/studies/generative-artificial-intelligence-capacity-study/occupation-data-on-ai-exposure

Acknowledgement of generative AI use

Generative AI (Anthropic’s Claude, Opus 4.8) was used to assist with R coding, data wrangling and visual-design decisions, in line with RMIT Library guidelines. All topic selection, data sourcing, analysis, narrative and final judgements are the author’s own.