AI Is Turning Office Demand Into a Forecasting Problem
Artificial intelligence is not only changing how companies work. It is changing how they measure the space they need to work in. A CFO report on Censuswide research for International Workplace Group shows a clear signal for property intelligence teams: office demand is becoming less predictable because workforce capacity, location strategy, and productivity models are all being recalculated at once.
Among 1,000 CEOs and CFOs in the U.S. and U.K., 60% said AI has made it harder to predict office space requirements over the next two years. That matters because commercial real estate planning has traditionally relied on relatively stable assumptions: headcount growth, lease duration, utilization rates, and geographic concentration. AI weakens each of those assumptions. If a company can automate parts of a workflow, redistribute teams, or support more asynchronous work, square footage becomes a variable rather than a fixed input.
The more important number is not the 60%. It is the 73% of executives who said technological change, including AI, has made their organization less willing to commit to long-term leases or traditional real estate solutions. This is a direct challenge to the underwriting logic behind office assets. If occupiers increasingly treat workspace as an adjustable operating layer, landlords and investors need better visibility into lease flexibility, renewal probability, desk utilization, and location elasticity.
The survey also found that 99.8% of executives are actively shifting real estate costs from fixed to more variable. That is close to a universal response, and it suggests the office market is not only reacting to remote work. It is reacting to uncertainty itself. Companies are trying to reduce exposure to long commitments while preserving optionality. In data terms, they are managing real estate as a dynamic portfolio rather than a static footprint.
The office is becoming a demand signal to be modeled, not a commitment to be assumed.
The location data is equally important. According to the report, 57% of executives are investing in hybrid workspace setups, 55% are establishing work locations closer to where employees live, and 52% are considering decentralized workspace models. This points to a more distributed map of demand. Instead of one central office absorbing most activity, companies may support smaller hubs, flexible memberships, satellite sites, and suburban or near-home workspaces.
For property analysts, that creates both risk and opportunity. Central business district demand may become more selective, with premium buildings still competing well while commodity office stock faces weaker long-term visibility. At the same time, flexible workspace operators and landlords with adaptable floorplates may gain value if they can prove utilization, not just occupancy. The metric to watch is no longer only leased square footage. It is how often space is used, by whom, and for what type of work.
The CFO article also points to a second research signal from JLL: only 15% of surveyed C-suite executives and corporate real estate leaders had moved beyond initial AI deployment to actively optimizing AI in real estate operations. Just 33% were actively modeling AI’s potential impact across locations and asset types. That gap is significant. Companies know AI will affect space demand, but many do not yet have the analytical infrastructure to quantify how.
The next advantage in office strategy will come from better scenario modeling. Occupiers should test how automation, hybrid attendance, hiring plans, and employee residential patterns change space needs by market. Investors should track flexible lease penetration, tenant exposure to AI-driven restructuring, and building-level utilization data. AI may not reduce office demand in a straight line, but it is making demand more fluid. The winners will be those who can measure that fluidity before it appears in vacancy data.
Source: CFO
