Well folks, it's been quite a week for AI in real estate — and honestly, you can't make this stuff up. We've got AI companies scooping up so much office space in Manhattan that landlords are popping champagne, a $50 million penthouse deal that nearly went sideways because both sides asked ChatGPT for advice (spoiler: it told them both to walk), and a new survey showing most Americans think AI is going to make housing even harder to afford. So yeah — AI is helping, hurting, and spooking the market all at the same time. Let's get into it.
1. AI Companies Are Filling Up NYC Office Towers — and Boosting Leasing by 152%
After five years of post-pandemic vacancy pain, New York City's office market is staging a comeback — and AI companies are leading the charge. According to Bloomberg and Savills data, AI firms added roughly 1 million square feet of office space across Manhattan in 2025, a 152% jump from the year prior. And they're not done: companies are actively hunting for an additional 1.4 million square feet. Names like Anthropic, OpenAI, Palantir, Harvey AI, and EliseAI are all expanding their NYC footprints, while legacy tech firms investing in AI capabilities added another 2.1 million square feet on top of that.
The result? New York posted its best year for office leasing since 2014, according to Savills. AI tenants are paying an average of about $88 per square foot — above the citywide average of $78 — and competing for flexible floor plans with good transit access. SL Green, one of the city's biggest landlords, expects over 900,000 square feet of leasing in Q1 2026 alone, a company record, with AI deals pushing two of its trophy towers to full occupancy.
Why It Matters: Here's the irony: AI is widely feared as a job killer that could hollow out office demand. But right now, it's doing the opposite — creating jobs, filling towers, and propping up commercial real estate in one of the most important markets on Earth. For CRE investors and brokers, the takeaway is clear: AI isn't just a tech story, it's a real estate story. And for residential agents in markets like NYC, more AI workers means more renters and buyers with tech-level incomes looking for housing. This boom has echoes of the dot-com era — the question is whether it has more staying power.
2. ChatGPT Nearly Blew Up a $50 Million Penthouse Deal — Ryan Serhant Had to Save It
On the flip side of the AI-in-real-estate coin, Netflix "Owning Manhattan" star Ryan Serhant shared a cautionary tale about a $50 million penthouse deal that nearly collapsed when both the buyer and seller independently consulted ChatGPT at the last minute. The seller asked if the price was too low. ChatGPT said yes. The buyer asked if they were overpaying. ChatGPT said yes. Both sides nearly walked.
The problem? ChatGPT could see comps and surface-level data, but it couldn't see the deal — the months of negotiation, the relationship dynamics, the timing pressures, or the trade-offs already baked in. As Serhant's colleague Ravi Kantha put it, AI models sometimes just tell you what you want to hear. Serhant eventually salvaged the deal after his social media post about the situation circulated back to both parties.
Why It Matters: This is the perfect companion story to the one above. AI is great at information problems — writing listing descriptions, running comps, organizing timelines. But it falls apart on judgment problems, where context, leverage, emotion, and relationship history matter. For agents working luxury or complex deals, this is the lane to own. The takeaway isn't anti-AI — Serhant's own firm is building AI tools. It's that knowing when NOT to use AI is becoming just as important as knowing how to use it.
3. Nearly 60% of Americans Think AI Will Make Housing Less Affordable
A new Redfin survey found that roughly 59% of U.S. residents believe AI advances will eliminate jobs and make it harder for people to afford homes. Only 30% think AI will strengthen the economy and improve affordability. The concern crosses party lines — 63% of Democrats and 57% of Republicans expressed worry about AI's impact on jobs and housing.
The fear isn't abstract. Some estimates suggest up to 30% of U.S. jobs could be displaced or significantly reshaped by AI, raising concerns about income stability that could ripple directly into housing demand. If prospective buyers feel uncertain about their job security, they're more likely to delay purchases — especially in a market where mortgage rates are still above 6%.
Why It Matters: Sentiment drives markets. Even if AI doesn't cause mass layoffs, the fear of it can suppress buyer confidence and slow transaction volume. Real estate professionals should be watching this sentiment closely — not just for macro forecasting, but for how they talk to clients. Buyers sitting on the fence because of AI anxiety need reassurance grounded in data, not dismissiveness. And for the industry at large, this is a reminder that AI's impact on real estate isn't just about tools and tech — it's about how people feel about their economic future.
That's it for this issue of AiRE Update. AI is already at the closing table — sometimes helping, sometimes hurting, and always reshaping how people think about buying and selling homes. Stay sharp out there.