A landmark federal government report has for the first time examined how artificial intelligence is reshaping Australia's workforce, identifying desk-based roles such as accountants, receptionists and marketing professionals as the most exposed to AI-driven disruption — while tradespeople and care workers remain on far more solid footing.

The report, which analyses jobs data across the four years since generative AI tools including ChatGPT entered mainstream use in 2022, finds that Australia's overall employment market has so far remained resilient in the face of rapidly advancing technology. However, it draws a clear dividing line between which workers face genuine risk and which are largely insulated.

Who Is Most Exposed to Artificial Intelligence Disruption?

The findings paint a picture defined largely by the nature of the work itself. Roles performed primarily at a desk — particularly those involving data processing, administration and knowledge-based tasks — are considered the most vulnerable to AI disruption. Accountants, receptionists and marketing professionals all feature prominently among the exposed occupations.

By contrast, hands-on roles requiring physical presence and human judgement are the least exposed. Plumbers, electricians, childcare workers and aged care workers are identified among those best placed to weather the AI storm, given the nature of their day-to-day tasks cannot easily be replicated by generative tools.

The report also reveals a notable gender and education divide. Workers in the most AI-exposed roles are more likely to be women and to hold university degrees. Men make up approximately 70 per cent of those employed in the safer, less-exposed categories, and fewer of those workers hold tertiary qualifications.

Australia's Jobs Market Is Holding Firm — For Now

Despite the risks identified, the overall picture remains cautiously positive. The report's chief economist and author noted that employment has continued to grow across the economy, including among young Australians who are considered among the most exposed demographic groups.

"We find that employment has continued to grow, it's continued to grow for young people as well who are thought to be most exposed and we're overall seeing ongoing resilience in the Australian jobs market," the report's author said.

That said, a meaningful gap is emerging between exposed and protected roles. Jobs considered most at risk of AI disruption recorded employment growth of 5.6 per cent between November 2022 and February 2026, compared with 9.5 per cent growth for those deemed least exposed — a difference of nearly four percentage points.

Not every high-exposure role is struggling. Software developer employment, for example, grew by a striking 25 per cent over the same period, bucking the broader trend for AI-exposed occupations.

Why Australia Isn't Just Catching Up

Some might argue the resilience seen so far simply reflects slow AI adoption — that the real disruption is yet to come as businesses catch up with the technology. The report's author pushed back on that interpretation, pointing out that Australia is already one of the world's biggest users of AI. He also noted that even the United States, which sits at the global frontier on this issue, is seeing its jobs market perform reasonably well.

Those two factors together suggest the current stability is not merely a lag effect, but may reflect a more durable adjustment dynamic than pessimists have feared.

Government Plans Ongoing Monitoring

Federal Employment Minister Amanda Rishworth welcomed the findings as encouraging, describing them as a useful baseline from which ongoing monitoring can be built.

"The good news is, we're not seeing a big shift or disruption," Rishworth said, adding that the government intends to use the framework established by this report to track developments over time.

The commitment to ongoing reporting signals that while policymakers are not sounding the alarm, they are watching closely — aware that the technology continues to evolve at pace and that today's resilience is no guarantee of tomorrow's stability.