Billionaire investor Mark Cuban is raising an eyebrow at all the AI hype, specifically the idea that it’s wiping out white-collar jobs.
“If AI is going to destroy white collar jobs first, shouldn’t we already be seeing occupancy declines in office buildings? Particularly in big cities where large employers are primarily based? Or am I missing something?” Cuban posted on X.
Turns out, he may actually be underestimating just how much office demand has already dropped. The U.S. office market is already facing a crisis, and the data backs it up.
Office Vacancies Hit Record Highs
According to Moody’s Analytics, the national office vacancy rate hit 20.7% in the second quarter of 2025, the highest ever recorded.
Cities that were once hotspots for white-collar jobs are seeing massive declines:
- San Francisco: 27.7%
- Downtown New York: 23%
- Charlotte: 23%
- Philadelphia (Center City): 20.4%
Remote and hybrid work have taken hold since the pandemic. Many companies now operate with smaller in-person teams, which means they need less space than before.
In big cities, entire office floors are going unused. Businesses are shrinking their leases or walking away from them altogether.
This shift isn’t just about worker preferences, it’s a cost-cutting move during a slower economy.
Landlords, Lenders, and Cities Feel the Pressure
Office landlords are stuck with falling rental income and rising debt costs. And the problem isn’t going away anytime soon.
Around $290 billion in loans backed by office buildings are due by the end of 2027.
Refinancing those loans is proving difficult, increasing the risk of defaults, especially for regional banks heavily exposed to commercial real estate.
Cities are also taking a hit. Lower property values mean shrinking tax revenues, which could result in cuts to public services like schools and transportation.
A drop in tax income can also result in delayed infrastructure improvements and less investment in city upkeep, which may make downtown cores even less attractive to businesses and workers.
What Comes Next: Conversions and Capital Shifts
Developers are trying to adapt. In some cities, older office towers are being converted into apartments.
Philadelphia, for example, has removed more than 1 million square feet of outdated office space from the market through residential conversions, which has helped stabilize local vacancy rates.
Nationwide, over 149 million square feet of office space is set for some form of conversion. But these projects aren’t simple.
Many downtown buildings weren’t designed for housing. They can be expensive to retrofit, especially when dealing with zoning rules, structural limitations, and outdated utility systems.
Not every location works well for residential use, either. Some are in areas with few amenities or limited public transportation, which can make them unattractive places to live.
Because of these challenges, investors are putting more of their money into other commercial sectors that are doing better.
Warehouses and logistics centers are still in demand, thanks to the rise of online shopping and the need for better supply chain infrastructure.
Even though vacancies there have ticked up to 7.1%, that’s still far better than the office sector and reflects ongoing demand for well-located industrial space.
What This Means for the Future of Work and Real Estate
Mark Cuban’s question hits on a real disconnect: if AI is supposed to be wiping out white-collar jobs, shouldn’t we be seeing less office demand?
The truth is, we already are. The office real estate market isn’t just slowing down; it’s fundamentally changing.
And that shift is bigger than AI. It’s about how and where people work in a post-pandemic world. Remote flexibility, economic uncertainty, and changing priorities are all part of the picture.
What we’re witnessing isn’t a short-term blip; it’s a reset. The traditional model of centralized office towers may no longer be the default.
Instead, companies are rethinking how much space they actually need, and whether they need it at all.
IMAGE CREDIT: “Mark Cuban” by Gage Skidmore, via Flickr. Licensed under CC BY-SA 2.0. Image adjusted for layout.