Entry-Level Employment Falls 19% in AI-Exposed Fields, Study Shows

Young workers face steepest AI-driven job losses

Employment among workers aged 22 to 25 in occupations most exposed to artificial intelligence has fallen 19 percent below their peers in less AI-affected fields, according to updated research from Stanford University economists. The gap has widened significantly from the 13 percent decline measured just one year earlier.

The August 2026 edition of "Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence" draws on anonymized payroll data from HR management company ADP to track employment trends across age groups and occupation types. The researchers rated each occupation's AI exposure using established labor market impact measures and real-world usage data from Anthropic's Claude model.

Automation versus augmentation

The employment effects vary sharply depending on how AI is being deployed. Jobs where AI primarily automates tasks—fully replacing human work—show the steepest declines in entry-level employment. Positions like accountants, auditors, and receptionists fall into this category.

By contrast, occupations where AI augments rather than replaces human expertise show much less impact on younger workers. Chief executives and registered nurses, for example, use AI tools to enhance their work but still require substantial human judgment and experience.

The data reveals the employment gap is driven primarily by reduced hiring of young workers rather than increased layoffs. Across the broader economy, overall employment levels in AI-impacted fields remain relatively stable, masking the concentrated effect on new entrants to the workforce.

Education offers mixed protection

The research identifies "codified knowledge"—formal, documented information that can be taught through textbooks and standardized procedures—as particularly vulnerable to AI disruption. Jobs requiring mainly this type of knowledge show slower entry-level employment growth compared to positions that rely on tacit knowledge gained through practice and mentorship.

Higher education provides some buffer against these trends. Occupations with larger shares of college graduates show more muted differences between AI-exposed and less-exposed fields. In jobs with fewer college graduates, the employment gap between high- and low-exposure occupations is more pronounced.

Since 2022, employment in the top 40 percent of AI-impacted jobs has fallen approximately 11 percent for workers aged 22 to 25. Meanwhile, employment in the 60 percent of jobs with least AI impact grew 10 percent for the same age group over the same period.

Why it matters

These findings suggest AI's labor market impact may be creating a generational divide rather than broad unemployment. Organizations maintaining current staffing levels while quietly reducing entry-level hiring could face talent pipeline problems within years. For workforce planning, the data indicates companies need strategies beyond simple headcount metrics to understand how AI adoption affects their ability to develop future leaders and maintain institutional knowledge transfer.

Lead researcher Erik Brynjolfsson told The Washington Post the trends point toward a labor market that "keeps its overall employment level while quietly closing the on-ramp for people starting their careers." The details were first reported by Ars Technica.