Data-centre gold rush: AI’s upside comes with a real inflation risk

Australia’s AI revolution is producing headlines about opportunity and innovation. Yet the same transformation is also concentrating demand in a narrow corner of the economy: data centres. Bloomberg Economics’ warning that an AI-driven data-centre boom could exceed 2 per cent of GDP and strain construction and energy markets should be taken seriously by policymakers and households alike. What ordinary Australians need to understand is simple. Rapid, concentrated investment in building and powering data centres is not like a spread-out wave of economic activity. It pulls on the same resources—construction labour, specialist materials and large amounts of electricity—that families, businesses and other industries need. When a sector absorbs capacity quickly, it can bid up wages and materials and push energy prices higher. That feeds through into inflation, raising costs for everyone from home owners paying mortgages to small businesses managing tight margins. For voters watching prices at the petrol pump, the supermarket and the hardware store, this is not an abstract economic debate. It is a reminder that the benefits of technological change do not arrive evenly. We can applaud the high-paid engineers and local contractors winning work, but we also have to manage the side-effects that make everyday living more expensive. The policy response should be pragmatic and focused. First, planning and approvals need to be rapid enough to support investment but rigorous enough to avoid bottlenecks that amplify inflationary pressures. If construction labour and materials are already stretched, governments should coordinate training programmes and supply-chain planning so data-centre development does not crowd out hospitals, schools and housing. Second, energy policy must reckon with surges in demand. Data centres are electricity-hungry, and unless their power needs are planned in concert with grid upgrades and clean-energy commitments, they will increase pressure on the system. That can mean higher bills for households and a larger reliance on fossil-fuel generation in the short term—an outcome at odds with emissions goals. Third, transparency over the economic footprint of large projects is crucial. Policymakers should publish assessments of how major developments are likely to affect local construction markets and energy grids. That will allow targeted mitigation—temporary labour mobility schemes, staged approvals or incentives for on-site renewables—rather than blunt, economy-wide policy shifts. There is a temptation to treat technological opportunity and inflation risk as separate debates. They are not. The promise of AI-driven prosperity must be managed so it does not arrive at the cost of higher living expenses for households already stretched by interest rates and housing costs. Australia can have both: a thriving AI and data-centre sector, and stable prices for consumers. But that outcome depends on clear planning, honest assessments of supply constraints, and policies that spread benefits while dampening the inflationary side-effects. If governments act now, they can turn a potential macroeconomic headache into an enduring comparative advantage.