Over the first nine months of this year, AI-focused firms leased more than 486,000 square feet of office space across Manhattan — already surpassing the 414,000 square feet leased by similar companies in all of 2024, according to Savills. The total is nearly double 2023’s figure, marking a sharp uptick in demand for workspace driven by smaller, well-funded, growth-stage tech companies.
While the city’s biggest tech tenants — Amazon, Salesforce, and OpenAI — continue to drive headlines with large expansions, brokers say the real story is happening below the surface. These firms may not command massive footprints yet, but collectively they are fueling an increasingly competitive market, according to BisNow.
Among the recent headliners: Salesforce added 71,000 square feet at 3 Bryant Park, OpenAI inked 90,000 square feet at the Puck Building last fall, and Sigma Computing took 64,000 square feet at One Madison Avenue this summer.
Still, much of the action is coming from startups leasing 5,000 square feet or less — often in Class-B buildings. Okada & Co., a New York-based real estate advisory firm, signed 10 leases with AI companies this summer alone.
Many of these tenants are signing short-term deals — often two to three years — but are backed by significant capital.
Stanford data shows that corporate investment in U.S. AI companies hit a record $252.3 billion last year. That momentum is now showing up in physical space.
There has been a rising demand from West Coast-based companies, particularly those rooted in Silicon Valley and the Stanford tech ecosystem, as they establish a growing presence in Manhattan.
Newmark is currently tracking 50 tech tenants actively seeking space in the city. Of those, 41% are looking to expand, while only a small minority are downsizing. Most deals are happening in Class-A buildings, but nearly half are still landing in more affordable Class-B properties — a balance between prestige and flexibility.
After years of muted activity from tech tenants during and immediately after the pandemic, it’s clear that AI companies are stepping in to fill the void.
