Now, the Australian data’s in. Women are far more likely than men to be in jobs at risk of being lost to AI automation.
After months of examining global reports on how AI will impact women’s workforce participation, this latest report examines the jobs most and least exposed here in Australia.
And it’s a government report, meaning we’re really hoping the Albanese Government is taking notice – as so far, Australia’s National AI plan has failed to seriously consider the gendered impacts of AI-related (or at least AI-blamed) job losses.
The Department of Employment and Workplace Relations (DEWR) report, AI and Employment in Australia, is the first time the Australian Government has outlined which occupations are most likely to be displaced by AI, and found that women and university-educated workers are most likely to be affected.
Almost seven in ten jobs in the “least exposed” quintile are held by men, compared with four in ten (43.7 per cent) in the “most exposed” quintile.
Meanwhile, jobs in the least-affected group grew by 9.5 per cent from November 2022, when ChatGPT was released to the mainstream, to February 2026, while jobs in the most highly exposed group grew by just 5.6 per cent. Some of this decline predates the 2022 period, but AI are now accelerating the tasks these roles are built on.
This latest data provides the local context for a well-documented pattern internationally including the ILO’s March 2026 brief, which found that women are far more likely than men to lose jobs to Gen AI in 88 per cent of countries analysed. They also found this is driven by women being more concentrated in clerical, administrative and business support roles.
So, should we be alarmed by this latest report? Is the record-high women’s workforce participation rate Australia posted in 2024 at risk, given the concentration of women-dominated occupations in the “most exposed” list?
The immediate ministerial response to today’s report includes attempts at reassurance and suggestions that job displacement isn’t as bad as some may have expected.
But that reassurance overlooks one key issue: the report period ends in February 2026, thereby missing what has arguably been the most transformative period for generative AI in history.
Optimistic reassurance?
Employment Minister Amanda Rishworth has sought to alleviate concerns, saying today that “this report shows labour market conditions remain strong by historical standards, youth outcomes have mostly held up, and occupation reshuffling has not accelerated.”
But the report’s February cutoff misses much of the 2026 period, which has fast accelerated the mainstreaming of the agentic layer of AI across employers, as those accessing generative AI moved from essentially getting drafting assistance to applying fully agentic systems capable of executing multi-step workflows with limited supervision. Things like invoice matching, data entry, expense auditing, and autonomous customer-service resolution and coding.
While the report does acknowledge its timeframe limitation, noting that the exposure scores are a snapshot and warning that “if AI model capabilities advance rapidly, these scores could become out of date,” it does not offer extensive predictions about future change, nor does it measure future improvements in technological capabilities. It also says that “AI has not yet been deployed by a sufficient number of employers” – which doesn’t seem to align with the rhetoric or with what shareholders would expect from large employers.
Also to note is that the report does not offer comprehensive analysis of specific demographic groups other than young people – where it devotes a chapter to describing youth as the “canaries in the coal mine”. Just one paragraph is dedicated to the finding that women are the group most concentrated in exposed occupations.
So I’m not convinced of the immediate reassurance being offered. Especially as we’re now regularly seeing headlines around large organisations cutting thousands of jobs, including in Australia. While in some cases, blaming job losses on AI is a convenient opportunity for companies to talk to their shareholders and investors, the gendered job losses are still happenning and the genered implications are concerning, in terms of which jobs are going, as well as which workers are the first to leave (for example, those with less opportunity to be “always on” and always present in an office).
The ten biggest occupations in the “most exposed” quintile are listed below, with additional information on the proportion of female workers in each, based on ABS and Jobs and Skills Australia data.
- General clerks (84% female)
- Retail managers (50% female)
- Software & Application Programmers (17% female)
- Accountants (52% female)
- Receptionists (95% female)
- Contract, Program & Project Administrators (55% female)
- Accounting Clerks (81% female)
- Purchasing & Supply Logistics Clerks (44% female)
- Advertising & Marketing Professionals (61% female)
- Checkout Operators & Office Cashiers (75% female)
We can see from the above that seven of the ten most exposed occupations are majority-female, with five being heavily female-dominated. While women are seriously underrepresented in the Software & Applications Programmers list, the report finds this occupation is has actually grown in recent years, with employment up 25% here since the ChatGPT launch.
Australia has worked hard to achieve record levels of women’s workforce participation and a historically low national gender pay gap.
The data indicates, without even including the past six months of AI rollouts and development, that we’re at risk of losing that progress. We need stronger plans for dealing with the gendered impacts of AI-enabled job displacement that go beyond building data centres to building and supporting Australia’s workforce.