Australia has a productivity problem that predates the arrival of ChatGPT, and the country’s economic managers have settled on an increasingly familiar answer: artificial intelligence. After years of flat output per hour worked, the argument coming out of Canberra and the corporate sector is that AI represents the single largest opportunity to reverse a slump that has quietly eroded living standards for the better part of two decades.
The framing was captured neatly in a recent piece headlined “Australia Bets on AI to Break Decades-Long Productivity Slump”, which lays out a wager now shared across government, the Reserve Bank and much of the business community. The bet is simple to state and fiendishly hard to win: that a wave of automation, better software and smarter decision-making will finally lift the productivity needle where wave after wave of reform has failed.
How bad is the slump, really
The numbers are sobering. Labour productivity growth in Australia has slowed to its weakest sustained pace in roughly sixty years, and the past decade has been especially grim. Productivity is not an abstract statistic beloved only by economists. It is the mechanism by which wages rise without inflation, by which the same number of workers can fund an ageing population, and by which the country pays for the services people expect. When it stalls, everything from real wages to the federal budget comes under strain.
The Productivity Commission, the government’s independent adviser on these questions, has spent the past year making the case that digital technology sits at the centre of any recovery. In its work on harnessing data and digital technology, the Commission argued that AI could add tens of billions of dollars to the national economy over the coming decade if adoption is handled sensibly and regulation does not smother experimentation before it starts. That analysis has become a reference point for a Treasurer keen to hang an economic reform agenda on something more concrete than good intentions.
The government’s pitch
Treasurer Jim Chalmers has repeatedly framed the productivity challenge as the defining economic task of the decade, and his economic reform roundtable last year put technology adoption near the top of the list. The government’s position is that Australia does not need to build the next frontier model to benefit. It needs to be a fast and confident adopter, threading AI through hospitals, mines, farms, warehouses and back offices so that the productivity gains show up in the places where most Australians actually work.
That distinction matters. Australia is a small player in the global race to build large language models, and few in Canberra pretend otherwise. The more realistic ambition is diffusion: getting proven tools into the hands of small and medium businesses that make up the bulk of the economy, and doing it faster than the last technology wave, when it took years for computers and the internet to translate into measurable output.
The sceptics have a case too
Not everyone is convinced the bet pays off on schedule, or at all. Economists point to what is sometimes called the productivity paradox, the awkward historical pattern in which transformative technologies take a long time to show up in the national accounts. The personal computer arrived in offices in the 1980s, yet the productivity boom did not materialise until the late 1990s, once workplaces had reorganised around the machines. There is no guarantee AI is any faster, and some evidence that early corporate deployments have delivered underwhelming returns.
Unions and worker advocates raise a different objection. Productivity gains that flow entirely to shareholders, while workers absorb the disruption, do not do much for the living standards the whole exercise is meant to protect. The Australian Council of Trade Unions has been blunt that AI adoption must come with guardrails on job security, retraining and the fair distribution of any windfall. There is also the uncomfortable reality that many of the most measurable “productivity” gains from AI so far involve cutting headcount rather than helping existing workers do more, which is a very different economic story.
Then there are the enablers that Australia does not yet have in place. AI at scale is hungry for electricity and data-centre capacity, and the country is already wrestling with an energy transition that is straining the grid. It is short of the specialised skills needed to deploy and govern these systems safely. And it lacks a settled regulatory framework, leaving businesses to guess at where the lines on privacy, liability and safety will eventually fall.
What it means for Australia
For Australian businesses and workers, the productivity bet is not a distant policy abstraction. It shapes whether wages can rise, whether the budget can fund aged care and the NDIS, and whether the country’s export industries stay competitive against economies that are adopting the same tools. Mining, agriculture and professional services, three sectors where Australia genuinely leads, are also three sectors where AI could compound existing advantages if the diffusion happens.
The risk is a two-speed outcome. Large firms with capital and technical talent capture the gains, while smaller operators and regional businesses fall further behind because they cannot afford the systems or the skills to run them. That is precisely the divide policymakers say they want to avoid, and it is the reason so much of the current debate centres on adoption support for small business rather than headline-grabbing investment in frontier research.
What is next
The next twelve months will test the wager. Expect further guidance from the Productivity Commission and Treasury on where AI adoption is delivering and where it is stalling, alongside continued argument over how hard to regulate a technology that is moving faster than any legislation can track. The government wants to be seen as pro-adoption without being reckless, a balance that is easy to describe and difficult to hold.
The honest position is that nobody yet knows whether AI will end Australia’s productivity slump or simply add a new chapter to the long history of technologies that promised more than they delivered on time. What is clear is that the country has decided to place the bet. The returns, if they come, will be measured not in press releases but in the slow, unglamorous statistic that quietly determines how prosperous the next generation of Australians gets to be.
Sources: JournalArta via GNews.
