New modelling shows the broader productivity impact of AI
Artificial intelligence could support up to 44,000 additional full-time equivalent jobs, according to new modelling, which maintains that the technology could be a potential solution to the country's long-running productivity problem.
EY-Parthenon's analysis projects that AI could support between 36,000 and 44,000 jobs, and add up to $116 billion to Australia's economy by 2036.
But Cherelle Murphy, EY Oceania's chief economist, pointed out that the jobs impact would be uneven across sectors.
"Construction is expected to see the largest increase in full-time jobs as AI adoption lifts demand for new capital, equipment, systems and infrastructure, including the data centres and supporting infrastructure needed to enable the technology," Murphy said.
Wholesale trade, retail trade, and transport and warehousing are also expected to benefit as higher productivity flows through to real wages and consumption.
Agriculture and mining, by contrast, are expected to shed roles as automation improves efficiency and reduces labour needs.
"The jobs story is not one of AI simply replacing workers across the economy. The modelling shows employment demand shifting toward sectors that benefit from stronger investment and household spending, while capital-intensive industries need fewer workers," Murphy said.
AI's broader productivity impact
Meanwhile, the analysis also projects that increased AI adoption could boost real GDP by between 2.6% and 3.2% over the next decade, driven by a 2.0% to 2.4% increase in multifactor productivity and between $31 billion and $38 billion in additional investment.
Murphy said the potential uplift went beyond a technology story.
"Productivity is the main driver of long-term economic expansion and improvements in living standards, but Australia's performance has been weak over the past decade," Murphy said.
"Labour productivity growth has averaged just 0.3% a year over the past 10 years, less than a quarter of the rate recorded in the previous decade. That is why the potential productivity uplift from AI matters, not just as a technology story, but as an economic growth story."
The findings arrive as Australia's productivity challenges have deepened.
Multifactor productivity, which measures how well labour and capital combine to produce outputs, declined 0.5% over the 12 months to June 2025, below the 20-year average of 0.4% growth per year, according to the Productivity Commission's Annual Productivity Bulletin 2026.
The Productivity Commission has previously pointed to AI as part of the solution. Its interim report on the productivity benefits of data and digital technology estimated productivity gains from AI would exceed 2.3% over the next decade, with labour productivity growth forecast at 4.3%. The EY-Parthenon modelling was based on assumptions aligned with those PC findings.
But converting AI adoption into actual productivity gains has proven difficult.
Microsoft's executive vice president for Copilot, Charles Lamanna, recently attributed lagging productivity to "institutional inertia" and employers' hesitation to take risks with the technology, despite 18 of ASX 20 firms and 66 federal government agencies already implementing Microsoft's Copilot.
Murphy said realising the economic benefits would depend on employers and government acting on workforce mobility and reskilling.
"Realising the economic benefits of AI will depend critically on workforce mobility and targeted reskilling. A key priority for employers and government should be helping workers move into the sectors and parts of businesses where demand is expected to grow," she said.
