What the Boom in AI Agents Means for Flexible Workspace

The flexible workspace industry has a tenant mix that looks roughly like this: software and IT, creative services, professional services (law, accounting, consulting), and a long tail of small businesses and solo professionals. Different operators index differently, but those four categories make up the bulk of the demand in most markets.

All four of those categories are about to be reshaped by AI agents. Not by chatbots. Not by co-pilots. By autonomous, multi-step systems that can do work that used to require an entry-level analyst, a junior dev, a paralegal, or a research associate.

So the question I keep getting from investors is the obvious one:

If AI eats the entry-level white-collar workforce, does it kill the demand for office space — and especially flexible workspace?

My answer is more interesting than yes or no. Let me walk through it.

The disruption is real, and it's hitting the bottom of the pyramid

Start with what the data says.

Per McKinsey's State of AI in 2025 report, 62% of organizations are now experimenting with AI agents, and 23% are scaling agentic systems within at least one business function. A median of 30% of survey respondents expect a decrease in workforce size in their business functions over the coming year specifically because of AI.

Goldman Sachs has been tracking the labor data and found that workers aged 22–25 in AI-exposed roles — software development, customer service, junior knowledge work — have seen a 16% drop in employment, while experienced workers in those same fields remain stable. Anthropic's Dario Amodei publicly predicted that AI could eliminate roughly 50% of white-collar entry-level positions within five years.

On the consulting side, the news is even sharper. McKinsey itself laid off roughly 200 technology and support staff in late 2025, with global managing partner Bob Sternfels signaling more non-client role reductions over the next two years. Bloomberg's reporting on the broader consulting industry found that firms like McKinsey, BCG, and Bain are quietly rewriting their entry-level hiring playbook.

HBR's coverage of this transition has framed it well: AI is not creating mass unemployment. It's creating a missing rung on the career ladder. Per the World Economic Forum's Future of Jobs Report, 92 million jobs may be displaced globally by 2030, offset by 170 million new roles — a net gain of 78 million. But the displacement and the reinstatement don't happen to the same people, in the same companies, in the same cities, on the same timeline.

The macro story is net positive. The micro story for a 23-year-old paralegal in a flex office in Sacramento is a lot harder.

Now let's translate that to demand for office space

This is where the analysis gets interesting for the flexible workspace industry. The standard fear is:

If half the entry-level workforce gets automated away, half the desks get automated away.

That's a clean argument. It's also wrong — for three reasons.

Reason 1: Senior workers don't take less space. They take more.

In an AI-agent world, the median knowledge worker is more senior, more experienced, and managing higher-leverage workflows than the median knowledge worker of 2019. Per McKinsey, the productivity gain from AI is concentrated at the top of the pyramid — senior workers using AI to do what used to be a team's worth of analytical work.

That doesn't shrink the office footprint per person. In some cases it grows. The economic surplus from AI is going to be captured disproportionately by the workers who learn to orchestrate agents. Those workers want better workspace, not less of it.

Reason 2: The entrepreneurial wave is coming.

Every prior wave of white-collar disruption created a corresponding wave of entrepreneurship. The dot-com bust gave us the post-2003 startup boom. The 2008 financial crisis gave us the 2009–2012 founder cohort that built Stripe, Airbnb, Slack, and Uber.

There is going to be a generation of laid-off mid-career professionals — consultants, marketing operators, junior partners, big law associates — who decide that the right response to "my firm is shrinking its associate class" is "I'm starting an AI-powered version of my firm." Those founders are the textbook flexible workspace tenant.

The HBR research on AI-era entrepreneurship is fairly consistent on this point: the cost of starting a knowledge-economy business has collapsed. A solo founder with three AI agents can credibly deliver the output of a five-person consultancy. That founder needs a private office, a meeting room, a mailing address, and a virtual receptionist.

That's exactly what we sell. 32% of Pacific Workplaces' revenue comes from those higher-margin services already — and that's the part of the business I expect to grow fastest in an AI-agent economy.

Reason 3: The quality bar is going up.

Here's the part I think most flex operators are still in denial about. The AI economy is going to be unkind to commodity flex space.

If your value proposition is "a desk in a shared room with WiFi," you're competing against the kitchen table, the local coffee shop, and a thousand undifferentiated coworking spaces. The AI-augmented professional doesn't want any of those. They want a private office with a door that closes, a quality meeting room they can take a client call from, fast internet, and an environment that signals seriousness to the people they're trying to win business from.

Per the Colliers 2026 Outlook, employers are "increasingly focused on balancing flexibility with operational efficiency, shaping demand for high-quality, well-located and amenity-rich office environments." That's the trade. The total addressable market for premium small-format office is expanding. The TAM for commodity hot-desking is not.

So does AI spell the end of the office?

No. But it does spell the end of a particular kind of office.

Big corporate floors full of associates doing the cognitive work that AI now does better, faster, and at near-zero marginal cost — that office is in real trouble. The flight-to-quality dynamic that CBRE has been calling out for three years is going to accelerate, not decelerate.

Smaller, higher-quality, service-rich offices — staffed by senior knowledge workers and AI-augmented solopreneurs — are going to be in more demand, not less.

The flexible workspace industry sits closer to the second category than the first. Whether any individual operator captures that demand depends on submarket selection, build quality, and whether they have the revenue diversification to weather the next 36 months of disruption.

Takeaways for operators

• Invest in product, not square footage. The trade for the next five years is not adding more desks. It's making the existing desks worth more to the right tenant.
• Index toward private offices and services. Hot-desking is a commodity. Private offices, meeting rooms, virtual offices, and concierge-style services are not.
• Build for the solopreneur and the 2–10 person team. That's where the demand creation is happening.

Takeaways for investors

• Operator quality is going to matter more than market beta. The next five years will separate the operators who actually understand product and unit economics from the ones who were just riding the post-COVID rebound.
• Revenue mix is a leading indicator. Operators where 25–35% of revenue comes from non-desk services have a much more defensible business model than operators who are purely desk-rental plays.
• Asset-light is the right side of the trade. If we're entering a period of demand reshuffling, operators who own real estate at the wrong basis are going to be the casualties. Operators who can move where the demand goes are going to be the winners.

AI agents are not going to kill the office. They're going to kill a specific kind of office and concentrate value in the kind that flexible workspace operators were already building.

Data Sources: McKinsey State of AI 2025; McKinsey AI in the Workplace 2025; Goldman Sachs Global Investment Research; World Economic Forum Future of Jobs Report 2025; HBR coverage of AI-era labor markets; Yale CELI Higher Education Summit survey; Bloomberg Businessweek (May 2026); Colliers 2026 Outlook Report; Federal Reserve Bank of New York; PwC Global AI Jobs Barometer 2025.

About Pacific Workplaces

Pacific Workplaces is a 23-year-old flexible workspace operator with 16 locations across California. Roughly 32% of revenue comes from higher-margin services — virtual offices, meeting rooms, and concierge support — that we believe will be the most defensible part of the business as the AI agent wave reshapes the demand mix. To follow Ben's work, subscribe to the PAC blog.