Australia’s central bank warns AI investment is adding inflation before productivity gains

Australia’s AI buildout is adding demand before delivering most of its expected productivity gains, the Reserve Bank of Australia warned in its November monetary policy assessment. Data-center investment is increasing inflationary pressure in an economy already facing capacity constraints, while most AI productivity benefits are likely to emerge beyond the next two years.

November’s forecasts put construction first

The bank’s November 3 outlook raised its forecast for economic growth over the year ahead, driven by stronger expected data-center investment. That is a near-term contribution to activity, distinct from the later productivity benefit the bank expects from AI. The investment can increase spending before the technology makes businesses more efficient.

Governor Michele Bullock described investment as much stronger than expected, including spending linked to data centers. The pressure comes from adding demand when the domestic economy has limited room to meet it. Separately, the bank’s monetary policy statement said AI-related investment was driving rapid price increases for technology-related goods, according to the Guardian.

This has added to inflationary pressures by adding to demand at a time when the domestic economy is already capacity constrained.

Michele Bullock, Reserve Bank of Australia governor, quoted by the Guardian

A$141 billion meets a constrained building sector

Figures from Deloitte Access Economics put the scale of that construction demand in perspective. Its assessment, reported by the Guardian on November 5, identified 37 data centers worth A$141 billion in Australia’s investment pipeline. The value was up 36% in a year, within a total investment pipeline of A$1.39 trillion.

Deloitte director Sheraan Underwood said the projects were competing directly with housing and infrastructure for workers and materials. High inflation left Australia with less room to absorb the investment surge. That identifies a concrete source of pressure: data-center builders need resources that other construction projects also need, rather than drawing on an unlimited supply.

Deloitte expected substantial construction gains, but a smaller direct economic payoff once the facilities were built. Much of the technology installed in them is imported, and operating the facilities requires relatively few workers. Its assessment separated the immediate building activity from the productivity dividend, warning that the latter could arrive much later.

Higher productivity would not guarantee lower rates

Asked when AI would start having a positive effect on inflation, Bullock answered that the timing was uncertain. She described a dramatic productivity improvement as conditional, not assured. Even if it happened, she said, faster growth could also raise the neutral interest rate—the rate that neither stimulates nor restrains the economy.

The warning does not make AI the sole explanation for Australia’s inflation outlook. The RBA also pointed to higher fuel and commodity costs linked to the Middle East conflict, tighter capacity than previously judged and increased inflation expectations. It forecast underlying inflation would remain above 3% until late 2027 and reach 2.5% around mid-2028.

Nor did the stronger data-center outlook translate into a lasting upgrade to the bank’s overall growth forecast. From mid-2027, it revised growth lower because of a higher assumed interest-rate path and weaker housing prospects. The forecast level of GDP at the end of 2028 was broadly unchanged from August. For now, the bank has incorporated more investment without assuming an immediate productivity-led escape from inflation.