AI is creating new challenges, opportunities and questions about the industry's future

Artificial intelligence is quickly becoming one of real estate's most important customers.

It's leasing offices, fueling data center development and helping drive some of the strongest pockets of demand the industry has seen since the pandemic. At the same time, it's raising uncomfortable questions about the future of the very jobs that have traditionally powered demand for office, residential and retail space.

In San Francisco and New York, AI companies have emerged as some of the most sought-after office tenants. In San Francisco, they accounted for nearly 40 percent of office leasing activity in the first quarter and now occupy roughly 7 million square feet of space, a figure projected to double by the end of the decade. As availability shrinks in some of the city's most desirable office districts, landlords are increasingly chasing AI firms in hopes of landing the next breakout success story.

New York is seeing a similar trend. AI tenants represented more than a third of tech-sector office demand in the first quarter, as venture-backed startups raced to secure space and expand headcount.

As expected, it's speeding things up. Brokers describe deals moving from initial tour to signed lease in a matter of weeks. Founders show up with ambitious growth plans, unconventional workplace expectations and, increasingly, AI-generated opinions about everything from floor plans to lease terms.

The culture accompanying the boom may prove just as consequential as the leasing itself.

Many AI firms are embracing intensive in-office work schedules, including the increasingly discussed "9-9-6" model of working from 9 a.m. to 9 p.m., six days a week. For landlords, that means buildings that are heavily utilized and tenants that place a premium on proximity. In New York, some office-to-residential developers are already targeting workers who would rather live a few blocks from the office than spend valuable time commuting.

Data centers have become one of the industry's fastest-growing asset classes, particularly in Texas, where billions of dollars of investment continue pouring into new projects. But communities are increasingly pushing back over concerns about power consumption, water usage and strain on local infrastructure.

In Illinois, lawmakers are debating whether the state's biometric privacy law is making it harder to attract the next generation of AI-related data center investment. Industry groups warn projects could migrate to neighboring states, while opponents argue privacy protections and environmental concerns should take precedence.

Yet even as AI creates demand across multiple property types, it's also forcing the industry to confront a longer-term, and in some cases, more sinister question: What happens if the technology eventually cuts the number of people needed to do office work?

Some in the industry argue AI will become another productivity tool that creates new categories of work and fuels economic growth. Others believe widespread adoption could slow hiring in office-heavy industries such as finance, law, accounting and consulting, reducing future space needs even if today's leasing activity remains strong.

A lot of that uncertainty is showing up in residential real estate.

Buyers and sellers are increasingly arriving armed with ChatGPT-generated pricing analyses, market forecasts and negotiating advice. Agents report spending more time correcting flawed assumptions, explaining local market nuances and responding to AI-generated questions that often overlook critical details such as off-market sales, concessions or local regulations.

Ironically, many agents are simultaneously embracing the technology themselves. Across markets such as South Florida, brokers are using AI tools to generate listing descriptions, marketing campaigns, videos and client communications, allowing them to automate routine tasks while focusing more attention on relationships and dealmaking.

That may ultimately be the most revealing lesson from real estate's early AI era.

The technology is proving remarkably effective at generating information, speeding up workflows and creating new sources of demand, but human judgment, for now, still plays a critical role.

By Mike Romano