AI-Exposed Majors See Lower Employment, Earnings for New Grads

Recent college graduates who majored in fields most vulnerable to artificial intelligence are experiencing measurably worse job outcomes than their peers, according to new federal research that quantifies AI's impact on entry-level employment.

A working paper released in September 2026 by the U.S. Census Bureau found that graduates from the most AI-exposed academic programs face both declining employment rates immediately after graduation and reduced starting salaries. The research represents some of the first concrete government data on how AI adoption is reshaping outcomes for young workers entering the labor force.

Documenting the AI employment gap

The Census Bureau analysis examined initial job placement and earnings for recent graduates, comparing outcomes across majors with varying levels of exposure to AI-driven automation and disruption. Programs deemed most vulnerable to AI showed statistically significant declines in both metrics.

The paper did not specify which majors fell into the high-exposure category or provide exact figures for the employment and earnings declines. However, the research establishes a measurable correlation between a program's AI vulnerability and its graduates' immediate labor market success.

Why it matters

This data arrives as colleges face mounting pressure to demonstrate return on investment and as prospective students increasingly weigh career outcomes in their enrollment decisions. If AI exposure becomes a documented factor in post-graduation success, universities may need to redesign curricula, adjust program offerings, or fundamentally rethink how they prepare students for an AI-transformed economy. The findings also suggest that the AI employment disruption many economists predicted is now materializing in measurable ways for the youngest workers.

Implications for higher education

The research could influence student decision-making around major selection, potentially accelerating shifts away from fields perceived as AI-vulnerable. Universities have already begun adapting programs to incorporate AI literacy and emphasize skills considered harder to automate, but this data may intensify those efforts.

The working paper's framing—comparing the current moment to "another Great Recession" for certain graduates—underscores the severity of the employment challenge. Unlike the 2008 financial crisis, which affected graduates broadly, this disruption appears concentrated in specific academic fields based on their susceptibility to AI.

The September 2026 release timing means the data likely captures graduates from the 2024 and 2025 academic years, a period when generative AI tools and other automation technologies saw rapid workplace adoption.

The findings were first reported by The Chronicle of Higher Education, which noted the paper provides concrete evidence of AI's reshaping effect on the job market for young graduates.