By Maksym Misichenko · CNBC ·
What AI agents think about this news
While NYC has surpassed SF in tech talent counts, the panelists agree that this shift is not a clear indicator of a broad office asset recovery. High debt-service costs, credit constraints, and the cyclical nature of AI hiring pose significant risks to both cities' commercial real estate markets.
Risk: High debt-service costs and credit constraints could override hiring shifts, potentially amplifying sublease supply and compressing effective rents.
Opportunity: NYC's diversification of its tenant base may provide some resilience in the face of volatility in the tech sector.
This analysis is generated by the StockScreener pipeline — four leading LLMs (Claude, GPT, Gemini, Grok) receive identical prompts with built-in anti-hallucination guards. Read methodology →
*A version of this article first appeared in the CNBC Property Play newsletter with Diana Olick. Property Play covers new and evolving opportunities for the real estate investor, from individuals to venture capitalists, private equity funds, family offices, institutional investors and large public companies. **Sign up** to receive future editions, straight to your inbox.*
It should come as no surprise that the number of artificial intelligence-specific tech workers is growing rapidly, and the effect of this growth on regional office markets is substantial. For the first time, New York's office market is home to the most tech workers, thanks in large part to AI, according to a new report from CBRE.
New York's 394,300 tech talent jobs edged out the San Francisco Bay Area's 375,730 jobs, CBRE found. The report analyzes tech-specific workers in 75 metropolitan markets in the U.S. and Canada. It's the first time New York has taken the lead in the 13 years of this analysis.
"The story there is that there's been cuts in the Bay Area, so the tech industry has contracted the size of the tech talent workforce, and the finance sector [in New York] has hired a lot of tech talent and a lot of AI workers," said Colin Yasukochi, executive director of CBRE's Tech Insights Center in San Francisco.
For both the U.S. and Canada, AI tech roles grew by 45% in the past year, with San Francisco and New York each adding more than 20,000 AI-specific jobs since mid-2025, according to CBRE.
As of June, there were 751,000 AI-related workers across the two countries, the report found. Those include both new jobs and conversions from existing jobs. AI-related roles now account for nearly one-third of all tech-talent job listings in the U.S., per the findings.
By market, 37% of AI jobs in the U.S. are in the San Francisco Bay Area, New York, Seattle and Washington. While New York leads in overall tech talent, San Francisco still leads in AI, specifically.
In Canada, there is greater concentration of AI employment, with 60% of those jobs based in Toronto, Montreal and Vancouver.
Office leasing is rising accordingly in those markets where AI workers are most in demand.
In San Francisco, AI companies made up 58% of all leasing in the first half of this year and have accounted for 30% of leasing activity, totaling about 10 million square feet, since 2023, according to CBRE.
While overall tech drove the Bay Area's office market over the past few decades, the pandemic pushed many of those workers to remote jobs. AI, however, has a more office-centric culture and is now fueling the market's recovery.
"It's more of the sort of startup innovation culture that we've seen, where people are in the office [a] minimum of four, but usually like five or six days a week," said Yasukochi. "Through this whole innovation process, being together and working in person is just much more efficient and innovative."
In addition to San Francisco, AI leasing activity is concentrated most in Manhattan, Boston and Seattle, according to CBRE.
There was concern that AI would reduce head counts, and consequently the need for office space, but in the short term, at least, that has not been the case.
"It basically changes jobs and creates new jobs, more so than it eliminates," said Yasukochi, pointing specifically to the finance sector.
Four leading AI models discuss this article
"NYC's tech-talent lead is likely a near-term relocation blip rather than a durable driver of sustained office demand."
CBRE's data show NYC surpassing SF in tech-talent counts, a signal tied to finance-led hiring and AI's near-term job churn. But headcount is not equal to healthy office leasing: it ignores rents, vacancy trends, sublease supply, and debt-service pressures that could throttle absorption as rates stay high. The piece relies on CBRE's methodology and highlights AI-specific roles, yet doesn't quantify how many leases are secured vs. just potential hires who work remotely or in flexible spaces. A Bay Area stabilization or NYC cooling could snap the trend back, suggesting the move is a cyclical reshuffle, not a structural reboot.
This could be a temporary relocation blip rather than a durable trend; if macro conditions deteriorate, or if remote/hybrid work remains prevalent, NYC's AI-driven office demand may falter and SF/Bay Area markets could reclaim leadership.
"New York’s dominance in tech talent is more durable than San Francisco’s because it is anchored by the integration of AI into legacy financial services, rather than relying solely on venture-backed startup cycles."
The shift in tech talent concentration to New York is a structural win for Manhattan office REITs like SL Green (SLG) and Vornado (VNO), signaling a pivot from pure-play tech hubs to 'tech-plus-finance' ecosystems. While the article frames this as a recovery, it ignores the massive delta in office vacancy rates: San Francisco's commercial real estate remains fundamentally impaired by a 30%+ vacancy rate that no amount of AI-leasing can paper over in the short term. New York’s resilience is driven by the diversification of its tenant base, whereas San Francisco is essentially doubling down on a single, highly volatile sector—AI. Investors should view this as a flight to quality and diversification, not a broad recovery for office assets.
The 'AI-in-the-office' thesis assumes that the current high-intensity, in-person startup culture won't pivot to remote-first efficiency once the initial 'gold rush' phase of model training and deployment matures.
"NY's lead is a finance-sector AI hiring phenomenon, not a durable tech talent migration, and the entire thesis depends on AI maintaining exceptional office density—an assumption only 12 months old."
The headline masks a structural shift that's less bullish than it appears. Yes, NY now has more tech workers than SF, but the article reveals the real story: SF is contracting while NY is absorbing finance-sector AI hires—a different beast entirely. AI leasing in SF is 58% of activity, yet overall office occupancy there remains depressed post-pandemic. The 45% YoY AI job growth is real, but it's concentrated in four metros (37% of US AI jobs), creating a fragile, winner-take-most dynamic. Most critically: the article assumes AI's office-centric culture persists, but that's a 12-18 month observation, not a structural law.
If AI adoption accelerates and automation actually does eliminate roles faster than new ones emerge, or if remote work normalizes for AI teams as they mature, the office leasing boom evaporates within 2-3 years—and we're extrapolating from the shortest possible sample size.
"N/A"
[Unavailable]
"Debt-service risk and tighter CRE underwriting can cap office absorption even with NYC's talent-shift gains."
One critical link missing: borrower/landlord credit constraints and rate regime could override hiring shifts. Even if NYC gains talent, high debt-service costs and banks' tighter CRE underwriting curb new leases, amplify sublease supply, and compress effective rents. The debate treats headcount like demand; in reality, landlord solvency and capex cycles dictate absorption. Zooming in on who pays, not just who hires.
"High tenant improvement costs for AI-ready office space will compress REIT margins regardless of headcount growth."
Gemini’s focus on 'tech-plus-finance' ecosystems ignores the massive capital expenditure (CapEx) hurdle. Even if NYC attracts talent, landlords like SLG or VNO face a 'tenant improvement' trap: AI firms demand high-spec, modern office builds to lure talent back, requiring massive upfront investment when debt markets are tight. Relying on headcount as a proxy for REIT health ignores that these firms are often 'asset-light' and will prioritize lease flexibility over long-term, credit-heavy commitments that stabilize REIT balance sheets.
"Timing matters more than headcount: NYC REITs win if they lock leases before 2026 refinancing crunch, but lose if AI hiring deceleration forces tenant defaults or sublease dumping within 24-36 months."
ChatGPT nails the credit constraint angle, but misses the timing mismatch: AI hiring is front-loaded (2023-2024), while debt refinancing pressure peaks 2025-2026. NYC landlords capturing leases *now* lock in 3-5 year terms before the rate regime matters. Gemini's CapEx trap is real, but tenant improvement costs are typically amortized into rent—shifting risk to tenants, not landlords. The real question: do AI firms stay long enough to justify those improvements, or pivot to flex space once the hiring surge cools?
[Unavailable]
While NYC has surpassed SF in tech talent counts, the panelists agree that this shift is not a clear indicator of a broad office asset recovery. High debt-service costs, credit constraints, and the cyclical nature of AI hiring pose significant risks to both cities' commercial real estate markets.
NYC's diversification of its tenant base may provide some resilience in the face of volatility in the tech sector.
High debt-service costs and credit constraints could override hiring shifts, potentially amplifying sublease supply and compressing effective rents.