Anthropic joins AI firms coming to New York City

AI Leasing Is Repricing New York’s Office Problem, But Not Solving Its Housing Problem

New York’s latest artificial intelligence office expansion cycle is more than a commercial real estate rebound story. It is a signal that the city’s post-pandemic office inventory is being sorted into three categories: trophy space that still commands institutional demand, obsolete buildings that may convert to housing, and lower-cost office assets that suddenly have a new tenant class.

Marketplace reported that Anthropic is expanding in New York City, joining AI and technology firms including OpenAI, Palantir, and EliseAI in seeking additional office space as they hire and scale. The strategic point for developers is not simply that AI companies are leasing. It is where they are leasing, what kind of buildings they can absorb, and how that demand changes the conversion math across Manhattan’s aging office stock.

For the strongest AI firms, the pattern is familiar. Well-capitalized companies want high-quality, amenitized, transit-accessible offices that help attract talent, host investors, and support sales growth. That reinforces the premium position of Class A and trophy assets in core locations. In a city where capital formation, finance, enterprise sales, legal services, and media all cluster at scale, office presence remains a strategic tool, not just a workplace decision.

The more interesting development signal sits below the trophy tier. Marketplace’s conversation with Bloomberg’s Natalie Wong points to rising demand from startups for cheaper Class B and Class C buildings, the same segment that has carried much of New York’s office distress. These properties have been central to the city’s office-to-residential conversion conversation, but conversion feasibility is never automatic. Floor plates, window access, plumbing cores, elevator layouts, zoning, location, acquisition basis, and construction costs determine whether a building can become housing or remains office by default.

That distinction matters for housing policy. New York needs residential supply, and office conversions can help, especially where older commercial buildings are structurally suited to reuse. But renewed office demand can compete with conversion momentum if owners see a credible path to lease-up without taking on the cost, entitlement complexity, and construction risk of residential redevelopment. A Class B building that looked like a conversion candidate in 2024 may look like a discounted growth office asset in 2026 if AI tenants can backfill space.

For planners, the question becomes more precise. The goal should not be to convert every weak office building. It should be to identify which assets are genuinely misaligned with long-term office demand and which corridors are better preserved for employment growth. A city that converts too broadly may lose flexible workspace needed by emerging sectors. A city that converts too cautiously will keep housing supply trapped behind outdated commercial assumptions.

Investors should also underwrite the time horizon carefully. AI leasing demand may be strong over the next three to five years, but office leases and capital improvements are long-duration bets. If AI ultimately reduces headcount in some sectors, today’s tenant expansion could become tomorrow’s demand uncertainty. That makes lease quality, tenant credit, building adaptability, and exit optionality more important than headline absorption.

The signal is clear: New York’s office market is no longer moving in one direction. AI demand is adding liquidity to parts of the inventory, while housing need continues to pressure the weakest and best-suited buildings toward conversion. Developers and city leaders should watch the spread between office rents, conversion costs, and residential values. That spread will decide which buildings remain workplaces, which become housing, and which continue to sit between both futures.

Source: Marketplace