AI-Driven Job Shifts in the U.S. 2025
Early Signs of Disruption and Adaptation
A St. Louis Fed study suggests the U.S. may be experiencing the initial stages of AI-driven job displacement, with a notable impact on sectors like computing and mathematics. These fields, with nearly 80% AI adoption, saw a 1.2% unemployment increase from 2022 to 2025, reflecting a 0.57 correlation between AI use and job losses. Entry-level workers face a 13% employment decline since November 2022, while experienced professionals in these areas gained 6% to 9% in jobs, according to recent payroll data.
Post-pandemic over-hiring by Big Tech, followed by layoffs—Meta’s 22% headcount reduction in 2023 and Google’s 2024 streamlining—complicates the picture. Some firms paid six-figure salaries to retain AI talent with little work, exposing inefficiencies. Meanwhile, industries with low AI adoption, such as personal services (stable employment) and legal/social services (18% adoption, declining unemployment), remain less affected.
Hiring trends show a pivot—58% of companies maintain recruitment levels but focus on AI roles, pausing non-AI positions, per Goldman Sachs. This shift, alongside economic factors like revised job data (258,000 lost in May-June revisions), raises questions about AI’s role versus broader uncertainties like tariff fears.