Demand for office space from companies focused on artificial intelligence surged in Manhattan in the first quarter of 2026, raising the prospect that annual leasing volumes could reach levels unseen since the dotcom era, according to reporting that cites data from Cushman & Wakefield and other industry sources.
Key figures and transactions
- In Q1 2026, AI companies leased roughly 93,000 square meters of office space in Manhattan, a volume that alone exceeds their entire leasing activity in 2025 (Cushman & Wakefield).
- The AI share of technology-sector office leases in Manhattan rose to 56 percent in the quarter, more than double the share for the whole of 2024 (Cushman & Wakefield).
- By contrast, Colliers data show that AI firms accounted for only about 2 percent of all Manhattan office leases last year, highlighting the pace of this year’s increase.
Notable deals include EliseAI’s lease of more than 10,000 square meters near Grand Central Terminal and the legal-tech startup Harvey’s planned move into 17,000 square meters in an office building overlooking Madison Square Park. Anthropic is also close to completing a lease for nearly 50,000 square meters in Hudson Square.
Growth patterns and office usage
Harvey has expanded rapidly: since signing its first Manhattan lease in early 2024, its New York headcount grew from 36 to more than 300. Smaller players are also scaling up: healthcare-focused Adonis initially leased 2,300 square meters at 3 World Trade Center for a 25-person team, and now employs roughly 85 people.
Many AI companies sign leases of five years or longer and emphasize in-person office presence, arguing that co-location of engineers and researchers accelerates innovation.
Dotcom parallels — and important differences
Observers frequently draw parallels to the 2000 dotcom bubble, when widespread lease terminations by unprofitable startups inflicted heavy losses on investors and real-estate stakeholders. The similarity is underscored by the fact that many AI companies today still prioritize market share and growth over near-term profitability and sometimes lease more space than they currently occupy.
However, a key difference is that leading AI firms are already generating substantial revenues and count large, stable corporations among their clients. Landlords have also grown more cautious: after lessons from the earlier tech crash, owners now scrutinize prospective tenants’ business plans and financials more thoroughly.
Risks and longer-term implications
The sector’s expansion carries potential risks for the office market. A recent CoStar survey found that 35 percent of respondents believe broader AI adoption will slightly reduce long-term office demand; six months earlier, that figure was 23 percent. The shift suggests growing concern among market participants about how AI might change office usage patterns.
Why this matters: AI-driven leasing is boosting Manhattan’s office market in the short term and could push 2026 toward a leasing peak not seen since 2000. At the same time, the combination of stronger corporate revenues in leading AI firms and landlords’ more conservative underwriting may mitigate the chance of a dotcom-style collapse, even as the industry’s long-term effect on office demand remains uncertain.
