PRINT AS PDF
The Federal Reserve’s Beige Book hints at a rapid shift in how Artificial Intelligence (AI) is beginning to influence labor market and business activity. Early observations from 2022–2023 focused on strong demand for AI-related IT services and selective hiring to support AI capabilities. By 2023, firms were starting to turn to automation and AI to cover understaffed or repetitive roles. By late 2025, the tone seemed to changed. Some employers were reducing headcounts, citing AI-driven productivity gains. Several firms reported AI replacing entry-level work or reducing the need for new hiring, suggesting that AI had moved from a complementary tool to a structural labor-substitution force in some areas.
Relevant Beige Book excerpts are below:
June 1, 2022: Professional and business services firms continued to report strong activity. Demand for IT-related services such as artificial intelligence, software services, and authentication services, was especially strong.
April 19, 2023: Contacts noted turning to automation to fill repetitive, understaffed roles, and some have begun to leverage the use of artificial intelligence in lieu of hiring for certain professional positions.
September 6, 2023: Hiring activity in the technology sector remained subdued aside from positions focusing on artificial intelligence.
October 15, 2025: In most Districts, more employers reported lowering head counts through layoffs and attrition, with contacts citing weaker demand, elevated economic uncertainty, and, in some cases, increased investment in artificial intelligence technologies.
November 26, 2025: A few firms noted that artificial intelligence replaced entry-level positions or made existing workers productive enough to curb new hiring.
Over the next five years, this emerging trajectory could evolve into three powerful scenarios: (1) Productivity-led restructuring, in which firms aggressively adopt AI to reduce labor costs and boost margins, premised on continued accuracy, reliability, and declining AI deployment costs. (2) Workforce polarization, where high-skill AI developers, managers, and integrators remain in demand, but entry-level, clerical, and routine knowledge work shrink, premised on employers believing AI can ramp without apprenticeship labor. (3) AI-enabled regional divergence, where metro areas with strong tech, research, and capital ecosystems accelerate economically while others lose jobs, premised on uneven access to AI infrastructure and talent.
A low-risk scenario would entail a “stalling out” of AI where adoption slows from a number of possible factors including unreliable outputs, cybersecurity incidents, rising regulatory barriers, chip or power shortages, and/or consumer distrust.
In sum, these references are among the first indications that AI will be impacting the economy in the years ahead. Occupations will face exposure ranging from high risk of automation to limited impact. Uncertainty about the precise boundaries remains fairly significant.
Nevertheless, a number of skills will likely prove beneficial. These skills concern human interaction (empathy, trust-building, etc.), systems thinking and judgment under uncertainty, oversight over AI (auditing, safety, ethics), interdisciplinary problem-solving, communication (persuasion and leadership), and field-based mechanical/technical/”hands-on” skills.
