Stanford University Survey Finds AI Is Impacting Entry-Level Jobs the Most

Among workers ages 22 to 25, entry-level employment has fallen significantly, particularly in occupations where AI is commonly used for automation.

Among workers ages 22 to 25, entry-level employment has fallen significantly, particularly in occupations where AI is commonly used for automation.

Brynjolfsson et al.

Entry-level employment trends are more difficult to interpret in occupations where AI is used to augment, rather than replace, human workers.

Entry-level employment trends are more difficult to interpret in occupations where AI is used to augment, rather than replace, human workers.

Brynjolfsson et al.

Among workers ages 22 to 25, entry-level employment has fallen significantly, particularly in occupations where AI is commonly used for automation.

Brynjolfsson et al.

Entry-level employment trends are more difficult to interpret in occupations where AI is used to augment, rather than replace, human workers.

Brynjolfsson et al.

How AI Is Affecting Entry-Level Jobs

Researchers also theorized that early-career workers may be especially vulnerable to the effects of artificial intelligence in jobs that depend heavily on “codified” knowledge. This refers to formal, standardized, and documented information that can be taught through education, textbooks, or written procedures.

That type of work differs from jobs in which AI primarily complements the “tacit” knowledge of experienced employees. Tacit knowledge is generally developed through hands-on practice, instruction, and repeated exposure to real-world situations, the researchers wrote.

To test this theory, the researchers used the level of formal education typically required for each occupation. They analyzed data from O*NET’s occupational database as an indicator of how strongly each job relies on codified knowledge.

The analysis found that occupations requiring more codified knowledge experienced slower employment growth for entry-level workers. By contrast, occupations that depended more heavily on tacit knowledge showed stronger employment growth among mid-career and senior employees.

Higher education may nevertheless provide some protection against the effects of AI on employment. The researchers found that differences between occupations with high and low AI exposure were more pronounced in fields with larger shares of college graduates. Occupations with fewer university graduates generally had lower exposure to AI, although some growing fields remain vulnerable to automation and workforce reductions.

In a recent interview with The Washington Post, lead researcher Erik Brynjolfsson said current trends could point toward a labor market in which some opportunities for new workers disappear while many jobs held by employees who entered the workforce before the AI boom remain in place.

“The entry-level impact we’re measuring is real, persistent and growing,” Brynjolfsson said. “I’m more worried than I thought about a labor market that is quietly closing the door for people starting their careers while maintaining overall employment levels.”