EMPLOYMENT STATISTICS

Labor-market statistics tracking employment, hiring, wages, participation, underemployment and other signals of artificial intelligence's impact on human labor.

What does the labor-market data show?

The Employment Statistics Tracker monitors labor-market indicators that can help reveal how artificial intelligence is changing demand for human labor.

We track employment, unemployment, hiring, job openings, wages, labor-force participation, underemployment and occupational trends, with particular attention to differences between workers and occupations with greater and lesser exposure to AI.

AI labor disruption may appear in the data before it appears in layoff announcements.

A worker who leaves and is not replaced may never appear in an AI layoff statistic. A graduate who is never hired may remain unemployed, accept work below their qualifications, move into another occupation or leave the labor force entirely. Each produces a different statistical signal.

Our standard is evidence. Labor-market changes are not attributed to AI merely because they occur during the AI transition. Wherever possible, we compare AI exposure, occupations, age groups, industries and time periods and distinguish observed statistical relationships from demonstrated causation.

Company-specific employment changes are tracked separately in AI Job Displacement, Hiring & Non-Backfills, and AI Job Creation.

AI EMPLOYMENT STATISTICS

The statistical record separates overall labor-market conditions from AI-specific signals. General labor-market weakness or strength is not, by itself, evidence of an AI employment effect.

U.S. LABOR-MARKET BASELINE

U.S. EMPLOYMENT SITUATION

AUGUST 2026

The U.S. labor market continued to add jobs in August, while unemployment remained at 4.1%.

Payroll employment
+162,000
Unemployment rate
4.1%
Labor-force participation
61.6%
Want a job, not in labor force
5.7 million
Information employment
−23,000 during August
Interpretation
The aggregate labor market does not show economy-wide employment collapse. Industry, occupation, age and AI-exposure differences therefore remain important for detecting possible AI-related changes beneath the national totals.
Primary source: U.S. Bureau of Labor Statistics — Employment Situation, August 2026

U.S. LABOR TURNOVER

JULY 2026

Millions of workers continue to leave jobs each month, creating a large potential channel for workforce adjustment through replacement hiring.

Job openings
7.3 million
Hires
5.1 million
Total separations
5.1 million
Quits
3.1 million
Layoffs & discharges
1.7 million
Interpretation
Large monthly worker flows mean employment can contract through reduced replacement hiring without requiring mass layoffs. JOLTS does not establish whether AI caused any particular hiring or replacement decision.
Primary source: U.S. Bureau of Labor Statistics — Job Openings and Labor Turnover, July 2026

AI-EXPOSURE SIGNALS

STANFORD DIGITAL ECONOMY LAB

AUGUST 2026

Young workers in highly AI-exposed occupations are increasingly falling behind their less-exposed peers.

Population
Workers ages 22–25
Measured gap
Employment in AI-exposed occupations stands 19% below where it would be if it had kept pace with less-exposed peers.
Primary mechanism
Reduced hiring rather than increased separations
Evidence type
Administrative payroll data covering millions of U.S. workers through June 2026
Evidence strength
DOCUMENTED ASSOCIATION
Interpretation
The results provide evidence of a growing employment divergence associated with AI exposure among young workers, but the researchers do not claim widespread economy-wide AI displacement or treat the result as definitive causal proof.
Source: Stanford Digital Economy Lab — Canaries in the Coal Mine?, revised August 12, 2026

U.S. CENSUS BUREAU RESEARCH

APRIL 2026

Early-career employment declined substantially in the industry-state groups with the greatest measured exposure to AI.

Population
Early-career workers ages 22–24
Measured change
Regression-adjusted employment in the most AI-exposed quintile declined 12% over the ten quarters following the introduction of ChatGPT, while less-exposed industries remained stable.
Primary mechanism
The decline in hires was the primary cause of the observed employment decline.
Evidence type
Matched employer-employee administrative data
Evidence strength
DOCUMENTED ASSOCIATION
Interpretation
The Census research independently identifies a substantial early-career employment divergence concentrated in more AI-exposed parts of the economy, with reduced hiring rather than layoffs as the principal measured mechanism. The result is evidence of an association and should not be interpreted as proof that AI caused every observed employment decline.
Source: U.S. Census Bureau — You're (not) Hired: Artificial Intelligence and Early Career Hiring in the Quarterly Workforce Indicators, CES-26-27