AI is already reshaping economies. Its supporting equity markets, driving investment cycles and beginning to change how work gets done. In a world navigating geopolitical fracture, the move to net zero and demographic ageing, AI has real potential to offset many of the global economy’s headwinds. The question for Australia is whether we extract its full value.
Investment in AI is accelerating at scale. Worldwide data centre spending is projected to exceed US$1 trillion annually by 2029. Australia is not missing out. Our strategic location, stable political environment and renewable power potential positions us favourably. Our estimates suggest the domestic data centre pipeline has jumped to roughly 6GW, or A$150 billion; capacity could more than triple by 2030. Despite this, vacancy rates are at record lows and utilisation is near full capacity. Private business investment rebounded sharply in late 2025, largely off the back of AI spending. The infrastructure build is well underway, and this pipeline is expected to provide strong ongoing support for business investment and GDP growth in Australia.
But infrastructure is only an enabler. The real prize on offer is sustained higher productivity growth. We estimate that economies which integrate AI deeply and widely could lift labour productivity by 0.8 to 1.0 percentage points per year. This would be a substantial uplift at a time when productivity growth has been missing in action. Compounded, those gains lift incomes, strengthen public finances and keep pressure off inflation and interest rates.
Higher productivity doesn’t necessarily mean fewer workers. More often, it means workers doing more valuable work. AI is likely to augment most jobs, not replace them, especially by taking over routine tasks. Early evidence shows it is already saving time, freeing up capacity and allowing firms to do more with the same workforce.
However, translating investment, and even adoption, into productivity gains is not automatic. Productivity gains require deep business transformation; new processes, new decisions and new skills. Adoption matters, but widespread institutional change and rewriting of existing business practices is what will move the needle. The countries that can achieve this more quickly are those that will realise the largest productivity gains.
Australia has a structural problem here. Despite strong data centre activity, the Productivity Commission puts Australia's likely AI productivity gain at just 0.4 percentage points per year, below many of our global peers. We grudgingly support this general view. There are long-standing features of our economy that hold us back from adopting new business technologies and practices.
These include relatively low rates of business dynamism, including slower firm entry and exit, and a long tail of smaller firms that have been slower to adopt new technologies.
As a result, technology adoption tends to be uneven across the economy. Larger, financially secure firms who have access to skilled labour and international best practice have historically been more dynamic and quicker to adapt. By contrast, SMEs, the public sector and the care economy have been much slower. Those sectors represent a growing share of our workforce and economy.
That makes AI primarily a leadership challenge, not a technology one. Business leaders cannot delegate this to IT departments. Extracting value requires genuine strategic intent, workforce investment and a fundamental willingness to rethink how organisations should function in an AI-enabled world.
Policy settings have an important role to play in whether new investment translates into economy-wide returns. Economic reforms that boost competition, cut red and green tape and improve business flexibility would materially improve the odds.
That starts with faster planning and approvals for major projects, reducing barriers to firm entry and expansion, improving skills and training systems, and making it easier for businesses, particularly SMEs, to adopt new technologies. These are not new ideas, but they become more urgent in an AI-driven economy where speed of adoption matters.
The opportunity is large, well-documented and time sensitive. Australia has secured a meaningful share of the investment phase. The harder task is converting that into lasting productivity gains. Whether Australia leads or lags is still a choice, but the window for making it won't stay open indefinitely.
