INTRODUCTION
Day One of the AFR Property Summit surfaced two distinct categories of change shaping Australian commercial real estate. The first is cyclical: interest rates and supply. Australia is the only G20 economy to have raised interest rates three times in 2026, after delivering three cuts in 2025. Ten-year bond yields have returned to levels last seen in 2011, and a residential property correction is now underway, which is forecast to reduce national house prices by up to 15% from their peak.
Despite this backdrop, commercial real estate fundamentals across most sectors remain robust, pointing to a market in which persistent scarcity of supply, rather than sentiment, is currently the dominant driver of value.
The second is structural: the reshaping of real estate by artificial intelligence; a foundational, general-purpose technology that could permanently change where and how real estate is used. Each of these operate on entirely different timelines and call for different responses.
CYCLICAL PRESSURES: RATES AND A SUPPLY CYCLE THAT WON'T EASE QUICKLY
FUNDAMENTALS HOLDING FIRM
Major retail centre landlords are reporting near zero vacancy rates, with visitation growth above 3% annually and continued cap rate compression in the convenience sub-sector. Prime office rents in supply-constrained precincts such as core Sydney CBD and Brisbane CBD are recording double-digit annual growth, underpinned by an absence of new stock through the end of the decade. Industrial vacancy has risen slightly but remains close to historic lows nationally. Across each of these sectors, rental growth is outpacing inflation, allowing net operating income to expand even as financing costs rise and has helped keep cap rate expansion more contained than the shift in bond yields alone would suggest.
COST INFLATION ERODING DEVELOPMENT FEASIBILITY
A common feature ties every asset class together within this cycle: an almost universal absence of new supply. Construction costs have risen more than 40% over five years, eroding development feasibility even where land and planning approvals are secured. Cushman & Wakefield's The Cost to Build: Office and The Cost to Build: Industrial research quantifies the extent to which rising input costs are now the binding constraint on new development feasibility, compounding land and planning hurdles.
This undersupply is not unique to Australia: comparable commentary on European and US commercial real estate points to supply sitting at all-time lows across most sectors, with limited buildable land compounding the effect of higher construction costs globally.
Data centres are the one exception to this trend amongst commercial real estate sectors. Global hyperscaler capital expenditure is tracking toward US$1 trillion in 2026. Cushman & Wakefield’s H1 APAC Data Centre Update shows that 1,372MW of new operational capacity were brought online in the first half of the year across the region, and an additional 7,103MW added to the development pipeline.
Despite this robust pipeline, demand continues to outpace supply. Sydney added 131MW over the same time period, but vacancy compressed to 2.2%. Deployment is being hampered by a distinctly cyclical constraint, land and power availability, rather than by demand. Site selection has shifted markedly from land-led to power-led criteria over the past decade, a reminder that even the most structurally-driven parts of the market remain subject, in the near term, to very cyclical constraints on the ground.
STRUCTURAL RESET: HOW AI PERMANENTLY CHANGES THE EQUATION
AI IS NOT A CYCLICAL PHENOMENON
Artificial intelligence sits in a different category to the pressures described above. It is not a feature of where the economy currently sits in its cycle; it is a permanent shift in the underlying operating environment for real estate, and it will continue to unfold irrespective of the short-term path of interest rates.
A SYSTEMIC, NOT A SECTOR, STORY
At the economic level, AI is expected to meaningfully lower the cost of knowledge work, lifting productivity and GDP growth while exerting disinflationary pressure and lowering risk premiums that should support real returns over the medium term. At the occupational level, it is set to reshape the activities carried out inside buildings and, by extension, which locations, specifications and building types remain optimal for tenants. Adoption within the Australian real estate industry remains comparatively nascent: available data puts serious AI use within the sector at roughly 11% of firms, against 24% in professional services and 38% in technology. This gap could close quickly as historically opaque and fragmented property data become more machine-readable. Cushman & Wakefield's AI Impact Australia Scenarios research explores this transition at the city level, as well as its sector-specific implications, in detail. AFR Property Summit 2026: Richard Pickering
THE QUALITY PREMIUM, IN EVERY SCENARIO
Across every scenario modelled in Cushman & Wakefield's AI Impact Australia Scenarios research, one finding holds consistently regardless of how the current rate cycle resolves: the premium commanded by quality stock increases. AI itself is not expected to generate a durable competitive advantage in isolation, placing a premium on real estate businesses building strong data foundations ahead of broader AI adoption. The sector-specific implications vary, but the direction is consistent:
- Office: demand bifurcates further between prime, amenity-rich space and commoditised stock.
- Retail: product discovery moves toward AI agents; the store's role shifts toward conversion and experience.
- Industrial: AI-enabled supply chain optimisation reshapes optimal locations and may support some re-shoring.
CONCLUSION
The cyclical pressures facing Australian commercial real estate today are genuine: a reversing rate cycle heightens repricing risk, albeit thus far offset by a shortage of new supply. These cyclical conditions call for investors to be prudent when selecting assets and deploying capital. Layered on top of them is a structural shift, the reshaping of real estate by AI, that will persist regardless of how the rate cycle eventually plays out. The structural transition AI represents calls for action on data foundations and asset quality, regardless of where the rate cycle currently sits.
