Despite widespread concerns that artificial intelligence could curtail the need for commercial real estate, a new report predicts gains from AI productivity will generate more demand for space.
The use of AI will boost U.S. property demand for industrial, office and retail space by about 12% in the next decade, adding about 330 million square feet, according to real estate services firm Cushman & Wakefield's baseline scenario forecast. The authors estimate the probability of the estimate being correct at 50%.
The forecast comes as AI's economic impact is "still very uncertain" and it's not yet clear how individual companies and broader industries will use AI, James Bohnaker, principal economist for Cushman & Wakefield and co-author of the report, told CoStar News.
"We are just starting to see some of the impacts of AI," Bohnaker said, adding that outside the massive investment in data center development, there's going to be a ripple effect in other commercial property types.
"Over a 10-year horizon, we see AI as being a net positive for the economy and businesses," he added.
The 50% probability of the prediction coming to pass reflects the rapid adoption and large degree of uncertainty over how AI will play out in coming years. Some economists predict AI will mean some businesses need fewer workers and therefore less office space, while others have suggested that an AI-induced boost in productivity lead companies to grow in new areas.
Cushman & Wakefield created four potential forecasting scenarios: its baseline of gradual AI adoption; rapid AI adoption; AI adoption falling short of expectations; and AI displacement in which more labor than expected is swapped for AI. The report measures total net absorption, the amount of tenant move-ins versus move-outs, between 2025 and 2035 from a pre- and post-AI lens.
The boost in real estate demand under the baseline scenario — the one with the highest probability rate — is largely tied to industrial properties, particularly for modern, power-intensive logistics and manufacturing space, the report found. Bohnaker said the industrial demand is exclusively for warehouses or distribution centers and does not include data centers, though these spaces may store or ship materials for data centers.
"AI will help with inventory management, and there will be new demand on the net to warehouse and manufacturing sectors," Bohnaker added.
For office, Bohnaker said he expects AI to accelerate a bifurcation of this real estate sector already underway, with demand shifting to high-quality, adaptable space. Cushman & Wakefield expects about 5 million square feet of new U.S. office space to be completed annually over the next few years — a sliver of what was once clocked at 50 million square feet a year.
"We are underbuilding the product the market needs most, which should put downward pressure on high-quality vacancy, even in a slower job growth environment," he added.