The Fed Sees AI Hiring Both Ways: Engineers Scarce, Entry Roles Weaker

Artificial intelligence is pulling parts of the labor market in opposite directions. In its September 2 national summary, the Federal Reserve found both positive and negative AI effects on labor demand, while skilled tradespeople and technical workers remained difficult to find.

The clearest example came from New York: contacts described AI engineers as scarce, while an employment agency saw weaker demand for some entry-level technology and administrative-support roles partly because of AI. The observations are qualitative accounts from regional contacts, not a representative estimate of employment gains or losses across the United States.

National hiring was subdued and uneven

Employment increased only very slightly during the period covered by the update. Three of the 12 Federal Reserve districts experienced modest gains, four recorded slight gains and five saw no change. Manufacturing, construction and some service industries generated the healthiest demand, while demand declined in retail and hospitality.

AI appeared within that uneven market as both a source of work and a reason for reduced labor demand. An independent Reuters account of the September 2 release also identified the positive and negative effects cited across districts and described the underlying material as qualitative economic information gathered from the regional banks.

The national evidence does not establish the number of jobs created or displaced by AI. It also does not show that every technical occupation faces a shortage. Its narrower finding is that employers’ demand and access to labor varied by occupation and industry, with AI contributing to movement in more than one direction.

New York puts specialist scarcity beside weaker junior demand

The New York district’s labor-market account pairs the two AI signals directly: highly skilled technology workers, especially AI engineers, remained difficult to source, while one employment agency encountered weaker demand for entry-level technology and administrative-support positions, partly because of AI.

That contrast appeared within an otherwise steady regional labor market. Overall demand held firm in a low-hire, low-fire environment. Finance workers remained in robust demand, while aerospace and semiconductor businesses contributed to stronger manufacturing hiring.

Head counts increased in manufacturing, construction, wholesale trade, transportation and information but declined in retail, personal services and education. Worker availability improved, and the regional evidence contained no signs of large-scale layoffs. The softer signal therefore concerned recruitment for particular kinds of entry-level work, not a documented wave of AI-driven dismissals.

The scope of the finding is important. “Engineers scarce” refers to sourcing difficulties for highly skilled technology workers, with AI engineers singled out. “Entry roles weaker” comes from one employment agency and covers technology and administrative support; AI was identified as one contributing factor rather than the sole cause.

The two signals reflect different kinds of work

A shortage of specialists can coexist with weaker junior recruitment because the positions serve different functions. Organizations introducing AI may seek experienced workers who can build, integrate, secure or supervise the technology while reducing recruitment for work that software can partly perform. The contact evidence establishes the simultaneous hiring signals, but it does not disclose the individual employers’ staffing decisions behind them.

Hiring demand can also weaken before total employment falls. Employers may open fewer positions, leave vacancies unfilled or change the experience required for a role without dismissing existing staff. The regional material does not distinguish among those mechanisms, so it cannot show how many openings disappeared or whether the shift will persist.

Fewer junior openings could eventually constrict the route through which workers gain the experience needed for specialist jobs. That is a possible consequence, not an outcome demonstrated by the current evidence. There are no figures here for affected vacancies, displaced workers, AI adoption rates or the duration of the change.

Contact evidence is not a national employment estimate

The AI findings should be read as anecdotal contact evidence, not as a representative measurement of the U.S. labor market. The publication compiles observations from businesses, employment agencies, community organizations, economists and other contacts across the 12 districts, using interviews, questionnaires and additional formal and informal channels.

The edition was prepared by the Federal Reserve Bank of Minneapolis from information collected through August 24. Its comments came from outside the Federal Reserve System and do not represent the views of Federal Reserve officials. In this context, “the Fed sees” means the central bank assembled and published conditions described by its regional contacts.

The evidence can identify an emerging tension: demand for advanced AI expertise may intensify even as some entry points into technology and office work weaken. It cannot yet determine the scale, geographic reach or durability of that pattern. Later regional updates and representative employment and vacancy data will be needed to establish whether the New York experience is spreading across employers and occupations.

Also read:

• AI Automates Tasks Before Whole Jobs—and Junior Hiring Is Already Weaker
• Top 8 Ways To Use Content Marketing to Grow Your Future
• The 73% Collapse: How AI Is Erasing Entry-Level Tech Jobs and Rewriting the Career Ladder