Graduating from college is meant to be an exciting milestone, a celebration of years of hard work. For many recent graduates, however, that excitement is increasingly paired with anxiety about the future. National trends in 2026 have suggested a stagnant labor market for job seekers, marked by slow employment growth, fewer job postings, and reduced hiring. And data from the Federal Reserve Bank of New York shows that unemployment rates among young graduates have been rising. As shown Figure 1 below, since 2022, unemployment rates among young graduates (22- to 27-year-olds with a bachelor’s degree) have surpassed those of all workers—a new trend not common until recently. As of March 2026, the unemployment rate for young graduates was 5.6%, compared with 4.2% for all workers, and 3.1% for all college graduates.
As a young graduate myself, I wanted to examine local labor market forces in Minnesota affecting entry-level employment. In my research, I came across a recent report from the Burning Glass Institute, “No Country for Young Grads.” In the report, researchers identify four interconnected forces behind the decline in entry-level opportunities: a post-pandemic shift toward leaner staffing and more risk-averse hiring, a growing supply of degree holders competing for a shrinking pool of professional roles, and the rise of artificial intelligence—both as a direct displacer of junior work and as an accelerant of the other trends.
On the AI front specifically, the researchers describe an “expertise upheaval”: because generative AI tools excel at foundational tasks typically assigned to junior employees—such as research, drafting, basic data analysis, and routine communication—employers may need fewer entry-level hires to complete the same work, while demand for experienced professionals who can direct and apply these tools remains steady or increases.
Since AI is an entirely new and unique factor young graduates are facing, this study inspired me to examine whether these “expertise upheaval” trends described by the Institute are playing out at the local level in Minnesota. To do this, I examined shifts in both entry-level and senior-level employer demand for various occupations based on their level of AI exposure.
To measure AI exposure, I used Yale Budget Lab’s average AI exposure scores for 23 major occupational groups. In their paper, “Labor Market AI Exposure: What Do We Know?” the authors synthesized six studies on AI exposure to create a single, composite AI exposure score. The score incorporates metrics such as the automation potential of skills and tasks, productivity gains for certain roles, and observed current AI usage. I then sorted the occupations by their AI exposure scores and split them into three equal groups representing low, medium, and high exposure (see Figure 2).
I then collected job posting data from Lightcast—a labor market analytics tool that aggregates millions of job postings alongside government data—to examine how employer demand has changed for entry-level occupations (i.e., those requiring 0–3 years of experience, or YOE) within each occupation group. Postings without a stated experience requirement were excluded.
Figure 3 below presents the share of entry-level job postings within each AI exposure category from 2020 to 2026. To smooth out seasonal fluctuations in hiring, the data are displayed as a 12-month moving average (MA). November 2022 is highlighted as AI’s “entry point” in the figure because it marks the public release of ChatGPT and the mainstream adoption of generative AI.
Since the introduction of AI, the share of entry-level job postings in high-exposure occupations has declined by 7.6%, while the share of low- and medium-exposure occupations has increased, rising by 2.4% and 5.2%, respectively. Put another way, in 2020, nearly 60% of all entry-level job postings were in fields such as business, sales, and other high-AI-exposure occupations. As of March 2026, that share had fallen to just 44.7%. These trends are broadly consistent with national findings from the Burning Glass Institute.
One unexpected finding is that medium-exposure occupations have grown more strongly than low-exposure occupations. This may reflect continued post-pandemic recovery in service-oriented occupations that fall within the medium-exposure category — including social services, protective services, personal care, and education — as well as ongoing growth in healthcare employment.
The high-AI-exposure group described above consists of eight major occupational categories. Table 1 below breaks out each category individually and shows the change in the share of total job postings since November 2022 for both entry-level roles (0–3 YOE) and senior-level roles (6+ YOE). Computer and mathematical, architecture and engineering, legal, and office and administrative support occupations show the largest declines in entry-level demand. More broadly, every occupation in the high-exposure group saw a decline in the share of entry-level postings and growth in the share of senior-level postings. This pattern aligns with the Burning Glass Institute’s “expertise upheaval” framework, which suggests that AI is more likely to disrupt entry-level work while having little effect on the demand for senior-level expertise.
Overall, my analysis reveals recent shifts in demand for entry-level occupations in Minnesota. Since 2022, the share of entry-level job postings in high-AI-exposure occupations has declined by nearly 8%, whereas senior-level roles in those same occupations have not been negatively impacted. While this analysis examines just one aspect of the labor market and does not control for broader economic and industry-specific factors, it seems that Minnesota’s entry-level labor market, particularly in high-AI-exposure fields, is experiencing major changes, and recent graduates may need to adapt their career strategy accordingly.