Particle.news

Download on the App Store

Stanford Study Finds Double-Digit Drop in Entry-Level Jobs for Young Workers in AI-Exposed Fields

Analysis of ADP payroll records links the decline to automation-focused uses of generative AI in fields like software development and customer service.

Gary Marcus on center stage during opening night at Web Summit Vancouver 2025.
Therese Waetcher (not pictured) using vibe coding to improve her Shopify storefront.

Overview

  • Researchers report about a 13% relative employment decline since late 2022 for workers ages 22–25 in occupations most exposed to generative AI.
  • The study draws on millions of U.S. payroll records from 2021 to mid‑2025 and finds the pattern persists after controls for COVID-era effects, remote work shifts, and firm-level shocks.
  • Older workers in the same occupations generally saw stable or rising employment, and pay trends show little change, indicating adjustment is occurring through fewer entry-level hires rather than wage cuts.
  • Employment declines are concentrated where AI automates routine tasks, while roles using AI to augment human work show stable or higher hiring for entry-level workers.
  • Coverage also notes a contested MIT claim that 95% of corporate GenAI projects fail, with critics highlighting methodological weaknesses and pointing to successful off-the-shelf deployments.