Will AI Displace Young Professionals? Key Findings from Stanford’s Labor Study

Study Overview

Researchers from Stanford’s Digital Economy Lab examined ADP payroll records up to July 2025. The dataset covers millions of employees at thousands of U.S. firms and contains three key variables for each worker: age, occupation, and an AI‑exposure index that quantifies the share of tasks in a job that can be performed by current generative‑AI systems.

Key Findings

High‑AI‑exposure occupations (e.g., software development, customer‑service) experienced a sharp decline in employment for newcomers aged 22‑25. Between the end of 2022 and July 2025, the employment rate for this age group fell by roughly 6 %, while workers older than 35 grew by 6‑9 % in the same occupations.

Overall labor market growth remained robust, but the growth trajectory for young workers stalled after 2022. In low‑AI‑exposure jobs (e.g., nursing assistants) employment continued to rise or stay stable across all age groups.

Automation vs. augmentation : In occupations where AI is deployed primarily for automation, entry‑level positions declined. In contrast, occupations where AI serves as an augmenting tool showed no significant employment loss for younger workers and, in some cases, modest gains.

Interpretation

The results suggest that AI more readily replaces procedural, textbook knowledge—skills that younger workers rely on—while tacit, experience‑based expertise held by older workers is harder to automate. Consequently, jobs with low returns on experience offer weaker protection for young entrants.

Reference

Paper: “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence”. URL: https://digitaleconomy.stanford.edu/wp-content/uploads/2025/08/Canaries_BrynjolfssonChandarChen.pdf

Figures

Code example

来源：机器之心
本文
约2000字
，建议阅读
5
分钟
那些年轻的职场人员到底会不会被淘汰？