The prevailing story about AI and the workplace has been one of contraction. Automation eliminates roles, headcount falls, and companies need less space to house a smaller workforce. It is a tidy narrative, it fits the anxiety of the moment, and it has shaped a great deal of commentary about where office demand is headed. It also does not match what the executives actually making these decisions say they are planning for.

JLL’s Future of Work Survey 2026, which gathered responses from more than 2,200 C-suite executives and corporate real estate leaders across 21 countries, found that most organizations expect to grow rather than shrink. “There were a lot of findings that are counter to the headlines we are seeing,” said Peter Miscovich, Executive Managing Director of Consulting at JLL. “Over 60 percent of companies we talked to expect to grow headcount.” The pattern underneath that number is one of targeted reductions in some functions paired with strategic expansion in others, with AI redesigning roles and shifting people toward higher-value work rather than removing them from the org chart. Notably, the same research found that skills gaps have overtaken budget as the primary constraint on real estate transformation for the first time in fifteen years, which suggests companies are more worried about finding the right people than about having too many.

Growing headcount does not translate cleanly into stable space demand, though, because the way that larger workforce operates is changing at the same time. The organizations doing serious scenario planning are landing on flexibility rather than commitment. “We are modeling scenarios about how the workforce will be impacted by AI and we think that means shorter leasing terms and the need for more flex space,” Miscovich said. That is a rational response to a genuine timing problem. Real estate decisions carry ten and fifteen year consequences, while AI’s effect on how work gets done is shifting on a timeline measured in quarters. Committing to a fifteen-year lease on a floor plate designed around current assumptions is a bet that those assumptions hold, and very few executives are willing to make it right now.

What may matter more than how much space companies take is what happens inside it, because some of the changes now being contemplated go to the basic physical assumptions of office design. The desk exists because people need somewhere to put a screen. Remove the screen from the equation and a considerable amount of workplace planning logic comes apart.

Companies are actively experimenting with voice as the primary way people interact with AI systems at work, which introduces acoustic problems that open-plan offices are catastrophically bad at solving. “We have clients that are working on audio interface in the office, which creates a lot of interesting challenges to how to create offices,” Miscovich said. “We will have to start thinking about things like audio shielding, individual speaker systems, and private zones.” An office where a meaningful share of work happens through spoken conversation with an AI system cannot function in the layout that has dominated workplace design for two decades. It requires something closer to a recording studio than a bullpen.

Voice is likely only an intermediate step. The interface question extends to hardware that does not sit on a surface at all, and the implications of that shift run deeper than acoustics. Miscovich expects the device to move onto the body. “I think wearables will eventually emerge as the new way we interact with AI,” he said. “When that happens, we have to shift in how we think about the need for screens and a desk to put it on.” Square footage per employee has been calculated around a workstation, a monitor, and the circulation space required to reach them. A workforce interacting with AI through wearable devices would require space planners to rebuild that math from an entirely different starting point, one where the fundamental unit is a person rather than a desk.

Running alongside all of this is a case for workplace quality that gets stronger rather than weaker as the technology advances. The physical elements that support cognitive performance, natural light, acoustics, thermal comfort, opportunities for movement, and dedicated focus zones, are functional requirements rather than perks, and the pressure to fund AI infrastructure has a tendency to push them down the priority list. Treating them as optional is a mistake, and the reason has to do with what working alongside AI actually demands of people.

“Working with AI requires a lot of emotional resilience, which can be taxing over time,” Miscovich said. “This will likely mean more importance placed on wellness as we deal with AI all day.” That is a less familiar argument than the productivity case for wellness design and possibly a more durable one. Supervising, correcting, and collaborating with AI systems is cognitively different work than executing tasks directly. It involves sustained judgment, continuous verification, and a kind of vigilance that is genuinely tiring in ways the industry has not yet learned to design around. An office built for that workforce may need to support recovery as much as it supports output.

None of this resolves the larger question of how much office space companies will need five years from now, and anyone claiming confidence about that answer should be treated skeptically. What is becoming clearer is the direction of the changes. Leases are getting shorter. Flexibility is becoming a requirement rather than a hedge. The physical assumptions underneath workplace design are being questioned in ways they have not been since the open plan arrived. And the companies furthest along are not the ones with the most certainty. They are the ones building the capacity to move quickly when the picture clarifies. The office is still part of the future of work. It is just not going to be the office we recognize.

The post AI Is Not Killing the Office But It Is Changing How We Use It appeared first on Propmodo.

​