TL;DR: Australia has not seen broad AI-driven job losses. A July 2026 government report found employment grew 5.6% in the most AI-exposed occupations since late 2022, compared with 9.5% in the least exposed. That gap is an early warning, not proof that AI caused job cuts.
Australians have heard plenty of confident claims about AI wiping out office jobs. Until now, most of those claims rested on company announcements, overseas research or forecasts about what software might do one day.
The Australian Government has now checked what has actually happened here. The Office of the Chief Economist compared official employment and vacancy data with Jobs and Skills Australia measures of generative AI exposure. The result is neither a jobs apocalypse nor a clean bill of health.
Has AI caused mass job losses in Australia?
No, the best current evidence does not show mass AI-driven job losses in Australia.
The Department of Employment and Workplace Relations published its AI and employment in Australia report on 8 July 2026. The department found that the labour market remained strong by historical standards, youth outcomes largely held up and workers were not switching occupations at an unusual rate.
Employment also rose across every AI-exposure group from November 2022 to February 2026. Growth was uneven: the most-exposed fifth of occupations grew by 5.6%, while the least-exposed fifth grew by 9.5%. That 3.9 percentage point gap deserves attention, but slower growth is not the same as jobs being lost. The department describes the evidence as an early sign of concentrated softening, not broad upheaval.
What did the Australian Government actually measure?
The report measured employment and hiring patterns, then tested whether more AI-exposed occupations had moved away from earlier trends.
The Office of the Chief Economist combined the ABS Labour Force Survey, Jobs and Skills Australia's Internet Vacancy Index and occupation exposure data from the 2025 Gen AI Capacity Study. The data runs to February 2026, even though the report was released in July.
The central model estimated that an occupation one standard deviation above average AI exposure had employment about 2% below its pre-ChatGPT trend by February 2026. That is a modelled difference from a counterfactual, not a count of people dismissed because of AI. The estimate also became statistically uncertain when researchers used two alternative exposure measures or excluded the COVID period. Those failed checks are a reason to watch the result, not turn it into a headline-sized certainty.
Which AI-exposed occupations are showing pressure?
Routine clerical and customer-service occupations show the clearest pressure, but the direction is not uniform.
The government report found call and contact centre employment below its November 2022 level. Local occupation data also records recent declines for call or contact centre workers and general clerks. Longer projections remain positive for both groups, while keyboard operators and filing and registry clerks have negative five-year growth estimates.
| Occupation | AI score out of 10 | Employed | One-year change | Ten-year growth |
|---|---|---|---|---|
| Keyboard operators | 8.0 | 53,100 | +2,500 | -3.5% |
| Telemarketers | 8.0 | 5,000 | 0 | +5.4% |
| Filing and registry clerks | 7.5 | 18,500 | +300 | -0.3% |
| Call or contact centre workers | 7.4 | 29,400 | -800 | +8.3% |
| General clerks | 7.0 | 286,600 | -3,900 | +19.0% |
The figures come from the site's current Jobs and Skills Australia and ABS dataset. They describe different periods and should be read as signals, not as proof that AI produced each change.
Why did programmer employment rise while AI coding improved?
Programmer employment rose because high AI exposure can increase demand for changed skills as well as automate routine tasks.
The government report found employment for software and applications programmers reached about 199,000 in February 2026, up 25% from November 2022. The site's current software and applications programmers profile records 195,400 workers, an AI score of 6.7 and official shortage status. Its ten-year employment projection is still positive at 26.7%.
Fresh hiring data points in the same mixed direction. PwC Australia's 2026 AI Jobs Barometer found Australian job advertisements seeking AI skills rose from 20,000 in 2024 to 41,000 in 2025. AI may reduce time spent writing routine code while raising demand for system design, testing, security, client judgement and oversight. A high exposure score means many tasks could change. It does not tell us whether the occupation will shrink.
Are young graduates already being pushed out?
No broad collapse is visible for young Australian graduates, though some entry routes may still be changing.
The July 2026 government report put young graduate unemployment at 5.4%, lower than at any point in the five years before COVID. Among employed graduates aged 20 to 24, the share working in degree-level occupations rose from 51.1% in November 2022 to 52.2% in November 2025. Those figures do not support a claim that AI has already shut graduates out of skilled work.
The risk sits inside jobs as much as between them. Drafting, basic coding, filing and routine analysis have traditionally helped junior staff learn. If software absorbs more of that work, employers may expect new starters to handle harder tasks sooner. PwC's 2026 analysis found new tasks in highly exposed occupations were 2.5 times more likely to require human-intensive abilities such as leadership, empathy and face-to-face interaction. That shift deserves monitoring even while headline employment stays steady.
What can the report not prove about AI job losses?
The report cannot prove that AI caused the slower growth seen in exposed occupations.
Australian job advertisements fell by at least 37% in every exposure group during the study period as the post-COVID hiring boom faded. Interest rates, migration, industry cycles, outsourcing and older forms of automation can all move occupation numbers. The government's models reduce some of that noise, but no statistical adjustment can identify every redundancy caused by a particular tool.
Counterevidence matters too. A CSIRO study published in April 2026 examined job advertisements from more than 4,000 Australian firms and found AI adopters posted 36% more non-AI jobs over time than similar non-adopters. That dataset mainly covered 2020 to 2023, before widespread generative AI use, so it cannot settle the current question either. The honest answer is that Australian evidence remains early and mixed.
What should Australian workers watch next?
Workers should watch task changes, hiring requirements and training access in their occupation, not a single national jobs number.
The ABS reported in June 2026 that 12% of Australian businesses used AI in 2024-25, up from 1% in 2021-22. Adoption ranged from 38% in information media and telecommunications to 1% in transport, postal and warehousing. Exposure is likely to arrive at different speeds across workplaces.
Training is another useful signal. The Hays Salary Guide research found 60% of surveyed Australian professionals used AI regularly, yet 78% of users had received no formal employer training. Jobs and Skills Australia's Australian Jobs 2026 says basic AI use, digital ability and human skills can help workers adjust as tasks change. Free, self-paced TAFE NSW AI microskills cover AI foundations, business use and responsible practice.
Frequently Asked Questions
Is AI taking jobs in Australia in 2026?
There is no evidence of broad AI-driven job losses across Australia. The Department of Employment and Workplace Relations found that employment still grew 5.6% in the most AI-exposed occupations from November 2022 to February 2026, although that was slower than the 9.5% growth in the least-exposed group.
Which Australian jobs show the most AI pressure?
Clerical and routine service roles show some of the clearest exposure. The site data gives keyboard operators and telemarketers an AI exposure score of 8.0 out of 10, filing and registry clerks 7.5, and call or contact centre workers 7.4.
Does a high AI exposure score mean a job will disappear?
No. Jobs and Skills Australia's 2026 report places only 4% of the workforce in occupations with high automation exposure, while 31% sit in high-augmentation occupations where AI is more likely to assist work. Exposure describes task potential, not a redundancy forecast.
Are Australian programmers already losing work to AI?
Programmers face high task exposure, but the employment evidence is mixed. The July 2026 government report found software and applications programmer employment rose 25% from November 2022, while the site data records an AI score of 6.7 and 195,400 workers.
Is AI hurting young Australian graduates?
The government report did not find a broad graduate collapse. Young graduate unemployment was 5.4%, and the share of employed graduates aged 20 to 24 working in degree-level occupations rose from 51.1% in November 2022 to 52.2% in November 2025.
How can Australians check their own job's AI exposure?
Will AI Take My Job scores 358 Australian occupations using Jobs and Skills Australia and ABS data. Check your occupation, then compare its employment, earnings and growth figures with the full rankings.
Key Takeaways
- Employment grew 5.6% in Australia's most AI-exposed occupations from November 2022 to February 2026, versus 9.5% in the least-exposed group.
- The government's central model found a 2% shortfall from trend for above-average exposure, but several checks made that estimate statistically uncertain.
- Software and applications programmer employment rose 25% after November 2022 despite the occupation carrying an AI score of 6.7 out of 10.
- Young graduate unemployment was 5.4%, while 52.2% of employed graduates aged 20 to 24 held degree-level jobs in November 2025.
AI has not produced a broad Australian jobs crash, but waiting for one national number to settle the issue will not help. Check the score for the work you actually do, compare occupations and pay attention to which tasks employers are changing first.