AI report warns 30 per cent of Australian workers could be affected within five years

More than 30 per cent of the workforce could be disrupted by the rapid uptake of artificial intelligence over the next five years without policy intervention, says a new report that warns that the sclerotic Australian economy was poorly equipped to deal with the next industrial shock.

It also warns that Australia risked being left behind all other OECD countries because of the structural nature of an economy geared heavily towards raw material exports and lacking high value industries.

Chief executive of the Australian Industry Group Innes Willox says successive governments have failed to grasp the consequences of the AI revolution, the pace of its adoption and the potential impact of on the workforce.

The modelling report conducted by the Social Policy Group, of which Mr Willox is also chair, presented a bleak view of the workforce impacts of AI, beginning with low-income jobs before quickly moving to high-income workers.

It warned of large scale redundancies by the end of the decade under a high adoption rate of AI technology, and assuming a business as usual approach, as the private sector sought to address the nation’s productivity crisis.

*Source: AI and the Great Retrenchment, Social Policy Group*

In a worse-case scenario, without any industry policy or regulatory intervention by government, up to 33 per cent of the workforce could experience disruption to their jobs by 2030.

A lesser scenario, based on economic modelling, suggested one in five workers could be affected.

Mr Willox said the report painted a worse-case scenario that was “bleak” but there was also scope for AI to be a job creator if the transition was managed well.

“The report is a warning of a worse-case scenario and a bleak view of what could potentially happen,” he said. “There is a lot that will happen to mitigate a lot of it. The report is a clarion call that change is coming, it’s coming quickly and it is going to have massive impacts on how we work, where we work, who we work for.

“It has to be a priority for whoever is next in government to grapple with this … we are the point of either we got on the bus or we will be run over by it.

“What history tells us is that every technical revolution has created more jobs than have been destroyed, and there is no reason to doubt this won’t the case with AI. But the challenge is to get us prepared and ready for it.”

Mr Willox warned of the potential for “dark factories” absent of workers, a phenomenon already occurring in countries such as South Korea. “What you see in businesses is a concern around the lack of productivity, which is steering more and more businesses towards digital solutions, and AI is a key part of that,” he said.

“This is the big challenge and I don’t think, politically, the change this will potentially bring has been grasped.

“That is the big take-out here and it goes to the heart of the productivity challenge. We have to get out skates on very quickly.

“The big challenge is for government to get the right setting in place to enable businesses and their workforces to adapt to this.”

The report warned of “tectonic” structural changes to the economy and workforce, with the AI transformation outpacing any previous technical revolution before it. “Historically, the diffusion of technological advancements was governed by physical capital investment, which created a natural brake on adoption,” the report said.

“The production, transportation, and integration of machinery required extensive supply chain co-ordination and significant financial outlay, while employees needed training to operate new systems. These constraints imposed delays that let labour markets and institutions adjust incrementally.

“AI obliterates this paradigm. With its intrinsic scalability, real-time digital distribution and ability to integrate into pre-existing workflows, AI accelerates adoption to a pace never before seen, leaving little room for traditional adjustment mechanisms.”

While the government recently released voluntary standards for the safety of AI, including 10 proposed guard rails, the report said the Australian economy more broadly was not geared to adapt to the rapid changes.

“While the first lay-offs begin in routine data entry, admin and retail, AI will proceed to replace many high-income, knowledge-based positions traditionally considered aspirational and secure,” the report said. “We already see this starting to unfold. The model finds Australia will be one of the worst-hit OECD countries.

“Australia’s economy is a composite of raw primary exports, low-yield domestic consumption services and high-income-generating professional services constructed for an Australian market.

“As a developed economy, Australia ranks among the lowest on the industrial complexity index, leaving us dependent on the knowledge economy and on retail and hospitality services to fuel employment.

“Without immediate intervention, Australia is at risk of large-scale job redundancies within the next five years.

“While lay-offs from AI will be global, Australia’s economic structure is uniquely ill-suited to absorb the shock or to capitalise on the potential productivity gains of artificial intelligence.

“In economies with robust industrial ecosystems and complex value chains, AI amplifies national wealth by increasing output and creating adjacent industries.

“By contrast, in an economy like Australia’s dominated by raw material exports, limited value-added production, and consumption-driven services, AI productivity gains are capped, translating primarily into lay-offs rather than growth.”

The chief executive of SPG, Carla Wilshire, said the peer-reviewed modelling, which aligns with international models, suggested that without intervention, the benefits of automation would bypass Australia and compound existing vulnerabilities.

“AI does not just replace tasks; it restructures economies. In nations with industrial complexity, AI-driven productivity reinforces high-value industries. In Australia, it exposes the absence of those industries,” Ms Wilshire said. “Our dependence on low-complexity exports limits our ability to reinvest productivity gains into industries that can sustain growth. Instead, the gains concentrate at the top, while the middle class, our economic stabiliser, contracts.

“Knowledge work, long considered the backbone of middle-class stability, is now at risk. AI is automating roles once thought immune, leaving displaced workers with few pathways to transition within Australia’s economic structure. The problem is not just job displacement; it is the lack of industrial ecosystems that can absorb the shock.

“Australia has the resources to compete but lacks the strategy to capitalise. We produce the critical minerals that power AI globally but export them as raw materials, forfeiting the value to nations that process and innovate with them.

“AI accelerates the pace of economic change, but it also offers an unprecedented opportunity. The decisions we make now will determine whether Australia leads this transition or is left behind, watching others set the terms of the future economy.”

Simon Benson is The Australian’s political editor. This article first appeared in the Technology section of The Australian.