Despite Rapid Uptake of AI, Businesses Say It’s Still More About Retraining Than Layoffs

4 September 2025

Limitless Visions - stock.adobe

The rising specter of artificial intelligence has been casting shadows across boardrooms and labor markets for years now. When businesses in the New York-Northern New Jersey region began reporting sharply increased use of AI, anxiety over mass job displacement quietly spread like wildfire. Yet when the Federal Reserve Bank of New York pulled back the curtain on its latest regional business survey, it revealed a far less dramatic picture. According to the Fed’s blog post published on September 4, AI has not so far triggered substantial layoffs instead, employers have leaned toward retraining existing staff.

The data underscores a swift shift in how businesses are integrating AI. In just one year, the share of service sector firms deploying AI surged from 25 percent to 40 percent. In manufacturing, adoption rose from 16 percent to 26 percent. Moreover, firms anticipate that this adoption will continue to grow, with nearly half of services companies and a third of manufacturers intending to roll out AI tools within the next six months.

Disturbingly, despite these adoption rates, layoffs tied directly to AI remain rare. Only a sliver of surveyed businesses some sources cite as few as 1 percent in services said they had laid off employees because of AI, and none of the manufacturers reported job cuts due to the technology. On the contrary, many companies responded by retraining workers or even expanding hiring to support AI initiatives.

This nuance matters greatly. It reflects a reality where AI is more frequently augmenting human work rather than replacing it. In many cases, AI is helping staff handle repetitive or information-heavy tasks, freeing them to focus on higher-value responsibilities demanding creativity or judgement. It’s not a wholesale replacement machine but rather, in many workplaces, a co-pilot for productivity.

Still, the calm may not last forever. The New York Fed cautions that as AI becomes further woven into operational workflows, firms may begin reducing hires or laying off employees systematically. The next six months could reveal whether retraining remains the norm or if automation shifts toward workforce reduction.

Plenty of voices in the broader conversation urge caution. Some observers worry that we are witnessing only the early, less disruptive phase of AI integration one where the technology is used to streamline, not eliminate, work. Critics anticipate that as systems grow more capable, AI could indeed target higher-level managerial and professional roles, once considered observationally insulated.

It is worth noting that the apparent absence of layoffs does not necessarily reflect stability throughout the job market. Some economists point to early signs of AI disrupting entry-level positions, where algorithmic systems can cheaply automate tasks that once required junior hires. Others note that even without outright layoffs, hiring freezes especially for younger or lower-paid roles could compress career progression and wage growth.

Indeed, one in-depth analysis from Barclays found that while AI exposure did not lead to mass unemployment, it correlated with slower wage growth especially in roles and sectors wherein AI can streamline routine tasks. The concern is that automation may stagnate earnings while adding complexity to the labor landscape.

Back at the Fed, economists emphasize that the findings represent a snapshot in time. The current impact of AI may be measured, but they urge careful observation as trends shift. Policymakers and business leaders are being pushed to prepare for a future where AI’s footprint in hiring decisions grows, and where a balance between technological efficiency and human inclusion must be struck.

In the end, what the report makes clear is that the narrative around AI and labor still has chapters to be written. The hype around job-killing algorithms oversimplifies a complex reality. For now, the most common answer to AI is not pink slips, it’s retraining, redeployment, and cautious experimentation. The real question now is whether that cautiousness will hold as automation accelerates. The Fed’s findings offer a rare window into how businesses are charting this terrain, one retraining module at a time.