---
id: "lib-2026-10-21-canaries-dashboard-employment-in-ai-exposed"
type: "library"
title: "Canaries Dashboard: Employment in AI-exposed occupations shrank in September"
lang: "EN"
org: "ADP Research; Stanford Digital Economy Lab"
source_type: "Other"
date: "2026-10-21"
week: "2026-W44"
url: "https://nosov.ai/library/lib-2026-10-21-canaries-dashboard-employment-in-ai-exposed/"
original_url: "https://www.adpresearch.com/research/canaries-dashboard-employment-in-ai-exposed-occupations-shrank-in-september"
tier: "B"
tier_name: "Useful, limited"
tier_reason: "A one-minute monthly release from ADP Research and the Stanford Digital Economy Lab on administrative payroll data; the figures are reliable, but the note gives no method detail or uncertainty and points to the dashboard for both."
summary_basis: "full-text"
original_minutes: "1"
tags: ["future of work", "youth employment", "ai exposure", "payroll data"]
---

# Canaries Dashboard: Employment in AI-exposed occupations shrank in September

Tier B (Useful, limited): A one-minute monthly release from ADP Research and the Stanford Digital Economy Lab on administrative payroll data; the figures are reliable, but the note gives no method detail or uncertainty and points to the dashboard for both.

Summary based on the full text.

## In short

The September 2026 reading of the Canaries Dashboard, published 21 October 2026 by ADP Research (author Nela Richardson) with the Stanford Digital Economy Lab. Employment in the occupations most exposed to AI fell 0.5 percent year on year while the least-exposed grew 1.0 percent. The gap is widest for the youngest workers: people aged 22 to 25 in high-exposure jobs saw a 3.7 percent fall, the 36th straight month of decline.

## Key numbers

- 0.5 percent: year-on-year fall in September employment in occupations most exposed to AI
- 1.0 percent: year-on-year growth in the least-exposed occupations
- 3.7 percent: fall for workers aged 22 to 25 in high-exposure jobs
- 36 straight months: run of year-on-year contraction for that group, starting October 2023
- 2.8 percent: fall for workers aged 26 to 30 in high-exposure jobs

## What it reports

- All ages: high-exposure occupations -0.5 percent, least-exposed +1.0 percent, September 2026 against September 2025.
- Ages 22 to 25: total employment -2.2 percent; high-exposure jobs -3.7 percent; least-exposed jobs +0.2 percent.
- The decline for young workers in high-exposure jobs began in October 2023, 12 months after ChatGPT's public release, and has continued for 36 months.
- Ages 26 to 30: high-exposure jobs -2.8 percent; least-exposed jobs +0.9 percent.

## How they got it

ADP payroll records for millions of workers in more than 730 occupations at tens of thousands of private US employers, observed monthly and linked to established measures of occupational AI exposure. The note itself carries no method detail beyond this; the dashboard holds the data and methodology.

## Limits

- A release note, not an analysis: no confidence intervals, controls or exposure measure named in the text.
- Covers private US employers on ADP payroll only.
- Exposure correlates with other forces (interest rates, sector cycles); the note does not separate them.
- No figures by gender, although the dashboard tracks gender.

## Why it matters for companies

This is a monthly series built on payroll records, so it is the closest thing to a running measurement of how entry-level hiring is shifting in AI-exposed work. Three years of decline for 22-to-25-year-olds in those jobs means fewer people are entering through the junior roles that used to train the next tier. Firms that cut junior intake in these occupations should plan where their future mid-level staff will learn the work.

Original: https://www.adpresearch.com/research/canaries-dashboard-employment-in-ai-exposed-occupations-shrank-in-september