The debate over whether AI will eliminate jobs has been raging for years. The early data now suggests a more nuanced reality: employment is holding steady, but hiring is slowing—and that distinction matters for business leaders making decisions about workforce strategy and location.

Over the past 18 months, a clear inflection point has emerged in professional and business services at the national level. Job openings, which peaked at more than 2.4 million in 2022 during the post-pandemic hiring surge, have fallen drasticaslly to roughly 1.2–1.3 million in 2025, with early 2026 data suggesting a further drop below 1.0 million. In total, hiring demand has declined by more than half from its peak.

Emerging evidence on AI adoption helps to explain this shift. Firms nationwide are scaling back hiring as productivity gains reduce the need for additional workers, particularly in what are considered routine or codifiable roles. Survey data from the Federal Reserve Bank of New York supports this dynamic; while few firms report AI-driven layoffs, roughly 12 percent of service firms have confirmed that they have already reduced hiring directly due to AI. Early research suggests that these effects are concentrated in entry-level positions, where tasks can be more easily automated.

Yet employment levels have continued to rise, and quit rates remain relatively stable. Data from the Bureau of Labor Statistics Current Employment Statistics survey shows that employment in legal services, a key component of the sector, has grown steadily in recent years. Total employment increased from approximately 1.14 million workers in 2019 to 1.21 million in January 2025, and to 1.23 million in January 2026, reinforcing that workforce levels continue to expand despite slower hiring.

In this context, the earliest impact of AI adoption is not widespread job loss, but a shift in how firms hire. Rather than expanding entry-level headcount, companies are becoming more selective, relying on productivity gains to reduce the need for routine roles. AI is increasingly automating codifiable, task-based work that has historically supported entry-level hiring, while complementing and boosting higher-skill roles tied to judgment, client interaction, and oversight. The result is a labor market that is not contracting, but rebalancing—away from broad-based hiring and toward more experienced talent.

Taken together, these trends suggest that while employment remains elevated, the pathway into these jobs is narrowing, particularly at the entry level. Broader labor market indicators reinforce this shift. Unemployment rates among workers with a bachelor’s degree or higher have remained stable over the past year, even declining slightly from approximately 3.0 percent in March 2025 to 2.8 percent in March 2026. The market for highly educated workers remains tight, even as it becomes more difficult to enter.

These patterns carry clear geographic implications. If AI’s early impact is concentrated in routine, back-office functions, its effects are unlikely to be evenly distributed. Pressure is likely to emerge first in suburban office markets, while central business districts—home to leadership, client-facing roles, and decision-making functions—remain relatively more resilient.

This dynamic is already visible in the New York metro area. Between the third quarter of 2024 and the third quarter of 2025, service sector employment declined in Hudson County, New Jersey, while increasing across the Hudson River in Manhattan. Real estate data points in the same direction. Leasing activity in Brooklyn declined sharply, with quarterly demand falling more than one-third and year-over-year volume down more than 60 percent, while leasing activity in Manhattan has remained relatively strong in comparison.

At the same time, Manhattan’s office market reflects a growing divide in expectations about the future of white-collar work. Commercial rents for Class A office space continue to reach record levels, suggesting that tenants still expect to maintain or expand headcount over the long term. Firms signing long-term leases are effectively betting that demand for office space will remain durable. This view is supported by steady levels of white-collar employment reported by the New York State Department of Labor.

Yet the market value of those same office assets has declined sharply. Office buildings in Manhattan are now valued below their mid-2020 levels, and shares of major office landlords have fallen significantly over the past 18 months, reflecting investor concern about long-term demand. This disconnect suggests that tenants and investors are operating under different assumptions about the impact of AI. Employers, closer to day-to-day labor market conditions, continue to invest in space, while investors are discounting long-term demand.

For site selectors and corporate decision-makers, the implications are significant. Slower hiring growth in white-collar sectors may reduce the pace of job creation typically associated with corporate expansions. At the same time, AI adoption and its implication for increased productivity of experienced talent in major labor markets is likely to relieve some of the recruitment constraints that have defined recent years.

The shift also calls for a reevaluation of workforce development strategy. Approaches that emphasize adaptability, reskilling, and long-term labor market resilience are likely to become more important than those focused solely on headcount growth. As AI reshapes demand for high-paid professional roles, job growth may increasingly occur in sectors such as healthcare, logistics, manufacturing, and retail industries, where local labor forces may be better positioned to meet demand.

In this environment, the central question is no longer simply how many jobs can be created, but how the nature of work is changing—and how companies and regions position themselves to adapt.