Has the rise of AI tools coincided with changes in employment levels in the United States? Specifically, have the employment levels of younger workers in AI-exposed fields been affected?
To find out, researchers at the Stanford Digital Economy Lab examined data from ADP, the largest payroll processing firm in the US, covering January 2021 through July 2025 (see the full report and methodology here).
The researchers found that although employment for other age groups continued to grow, early-career workers in the most AI-exposed occupations experienced a 13% relative decline in employment since the widespread introduction of generative AI tools in October 2022.
For example, employment for software developers age 22-25 declined nearly 20% from its peak in late 2022.
The researchers segmented the US employment data into different quintiles of AI exposure and found that younger workers in fields like marketing and sales managers (fourth quintile; high AI exposure) experienced declines in employment, whereas those in fields like health aides (first quintile; low AI exposure) experienced increased employment in the same time period relative to other age groups.
The potential impact on employment in AI-exposed fields seems to be specific to early-career workers. In fact, the researchers found that although employment for 22-25-year-olds in the highest two exposure quintiles declined 6% between late 2022 and July 2025, employment for workers age 35-49 actually grew 9%.
About the research: The report was based on data from ADP, the largest payroll processing firm in the US, covering January 2021 through July 2025.
