Studies find declining employment opportunities for early-career workers in AI-exposed fields

Studies have recently found that young workers in AI-exposed fields are seeing a decline in employment opportunities. Several factors are at play.

A recent study by Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen for the Stanford Digital Economy Lab shows that workers in AI-exposed industries between the ages of 22 and 25 are seeing fewer employment opportunities than workers in less-exposed occupations. More experienced (older) workers in the same AI-exposed industries are not experiencing the same decline in opportunities and are either remaining stable in their careers or being presented with growth potential.

According to data from the study, which was reviewed and reported on by the Federal Reserve Bank of Dallas, employment for young workers in AI-exposed occupations fell from around 16.4% in 2022—when ChatGPT was initially released—to 15.5% in 2025. The report states that “This pattern suggests that the effect so far reflects changing labor market dynamics rather than widespread displacement.” Essentially, this means that a variety of factors are contributing to the decline in jobs for young workers in AI-exposed industries—the introduction and increased usage of ChatGPT is not solely to blame, though it is part of the issue.

Fewer young workers appear to be entering these jobs, which is also affecting the drop. Economists have noted that the decline does not seem to be driven by lower job-finding rates or layoffs from those who are already employed in AI-exposed industries. They have warned, however, that even though 22-25-year-old workers make up only about 9% of the total labor force, their decreased employment is shifting labor patterns. It’s unclear what the long-term effects of this may be.

The authors of the Federal Reserve Bank of Dallas report, Tyler Atkinson and Shane Yamco, are advising people to remember that there’s a complex correlation between employment and AI exposure, and that demographic shifts and education levels must also be considered when studying this decline. The primary thing they’re trying to convey through this information is that the long-term impacts of AI on the labor industry are currently unknowable; patterns will naturally reveal themselves over time as we move into an increasingly technology-driven landscape.

How the information was collected & what industries are being impacted

Per reporting from the Federal Reserve Bank of Dallas, there are lower employment rates for 22 to 25-year-old workers in industries with higher AI exposure due to several factors, including “a decline in people transitioning directly from out of the workforce into employment.” The study analyzed data from ADP, which is the largest payroll software provider in the country (approximately 25 million US workers are enrolled in ADP’s payroll services). This information, coupled with an AI exposure score developed by Stanford University researchers, helped measure whether certain employment tasks were being affected by the widespread use of AI.

From there, 12-month moving averages for labor and employment transition rates were considered to ensure the results were as accurate as possible. Researchers found that jobs with the least AI exposure were as follows:

• Freight, stock, and material movers
• Cashiers
• Laborers
• Building cleaners
• Janitors

Jobs with moderate AI exposure included:

• Elementary and middle school teachers
• Sales workers
• Retail salespersons
• Truck drivers

And the jobs with the highest AI exposure (AKA the areas where young workers are seeing the biggest decline in employment opportunities) are:

• Administrative assistants and secretaries
• First-line supervisors of retail sales workers
• Customer service representatives

Fort Worth Inc. noted that it’s important to have access to this type of information, as it allows for a more complete overview of AI and how it may negatively affect labor markets and long-term economic growth.