AI is the most recent set-back to office occupancy - by Daniel Calano

Corporate leaders are talking about Artificial Intelligence (AI) at every corporate retreat, and so are their real estate advisors…with good reasons. As AI reshapes staffing models, it is already influencing how much space companies, where they want it, and how it should be designed. 

The result is not a simple more or less jobs, but a re-balancing across office markets.

From a staffing perspective, AI is driving three main shifts:
1. Automation of routine;
2. Augmentation of knowledge workers, enabling higher outputs; and
3.Creation of new roles in data engineering and AI governance.

These changes flow straight into space planning. Way back, changes made came out of desperation (COVID). Then came work-from home; next zoom…now hybrid combinations, pushing companies away from one desk per person employee model.

 Currently, CBRE’s office occupier and workplace technology research shows that only about 40% of companies maintain a 1:1 employee-to-desk ratio, down from around 56% in the prior year, as organizations lean into shared seating and “occupancy management.” JLL’s work research and surveys similarly show that hybrid is a major driver of work-place design, even as some workers want to return to office life.

If AI allows companies to grow revenues with fewer office staff, it’s clear that there would be a drag on office space demand. We are already seeing higher vacancies in many markets. That said, there is also some irony in that AI companies themselves are producing successful AI and occupying more space to research and produce their work. For example, in the San Francisco Bay area, JLL researched that six new leases of about 280,000 s/f in 2004 have increased by 41 new leases covering about 2.4 million s/f in 2025. Similarly, according to NAIOP’s Office space Demand Forecast reports that net office absorption turned positive in the second and third quarters of 2024, and projects modest demand growth in 2025/26.

Also ironically, corporate occupiers in areas like Boston and Cambridge are using ever improving AI to improve building operations that optimize HVAC energy use, better maintenance, space planning, and so on. The pluses and minuses are yet to be quantified, and we will not know net outcomes for a while. The question of the moment is whether AI is a job improver, or an employee subtractor…or both. In any case, my bet is that buildings will be modified and/or built to create a good balance.

Daniel Calano, CRE, is managing partner and principal of Prospectus, LLC, Cambridge, Mass.