The Silent Crisis: How AI is Reshaping Entry-Level Work

Artificial intelligence has not yet led to widespread mass unemployment, with aggregate employment in developed nations remaining largely stable. However, a concerning shift is occurring beneath the surface: a quiet erosion of the foundational first rung of the career ladder. Recent assessments, while showing limited evidence of AI significantly altering headline employment figures, point to a troubling trend in early-career hiring. The most compelling evidence of this shift is manifesting precisely where it would be expected first: among early-career professionals. A working paper from the Stanford Digital Economy Lab, released in November 2025, revealed a 16% relative decline in employment for workers aged 22 to 25 in occupations highly exposed to AI, following the widespread adoption of generative AI. This decline was observed even after accounting for other factors influencing hiring decisions. Interestingly, more experienced workers in the same fields did not experience a similar downturn, nor did entry-level jobs with low AI exposure. This suggests the concern is specific to early-career roles susceptible to AI automation, such as software developers, customer service representatives, computer programmers, and information systems managers, where AI can substitute for junior tasks. This is not an insignificant signal. It indicates that businesses may be leveraging AI to perform the junior-level tasks through which individuals traditionally gain their initial professional foothold. This trend is exacerbated by a broader softening of the labor market for recent graduates. The Federal Reserve Bank of New York reported that in Q4 2025, the unemployment rate for recent college graduates climbed to 5.6%, and the underemployment rate reached 42.5% – its highest since the COVID-19 pandemic. While no single statistic can definitively attribute this solely to AI, and general hiring has slowed post-pandemic, it would be a mistake to disregard AI's potential role in accelerating an already challenging transition from education to employment. Behind these statistics lies significant personal distress. Recent graduates frequently submit hundreds of applications before securing an offer, with surveys consistently highlighting elevated levels of anxiety, financial precarity, and burnout among young job seekers. If AI quietly closes the door on typical entry-level positions, individuals will face delayed independence, postponed family formation, and the disheartening sense that their initial professional endeavors have been rejected. Furthermore, entry-level jobs are crucial components of the economy's informal training system, where junior staff learn practical skills, judgment, and the nuances of real-world operations. If AI absorbs tasks like drafting, triage, coding, and administrative preparation, firms might gain short-term efficiency, but society risks becoming less capable in the long run. The traditional advice to "learn to code," which underpinned over a decade of educational initiatives, is increasingly outdated. That premise, based on coding being a stable, scalable skill for middle-class jobs, no longer holds true, as AI now excels at routine coding, pattern reproduction, and debugging predictable errors – precisely the tasks these programs focused on. Instead, the relevant skills now revolve around supervising AI systems, understanding their outputs, and combining AI capabilities with human expertise. Educational institutions must adapt by embedding AI literacy, data literacy, prompt-based workflow skills, verification skills, and domain judgment into all ordinary degrees. To prepare young people for this AI-augmented workforce, a multi-faceted approach is essential. Schools should emphasize practical, real-world experience through paid co-ops, apprenticeships, and employer-linked projects, allowing students to develop critical judgment before graduation. Governments must incentivize businesses to hire and train early-career workers through targeted tax credits, wage subsidies, and training grants. Businesses, in turn, need to recognize the long-term value of cultivating an AI-experienced workforce, a process that inherently starts with robust entry-level opportunities. Students themselves bear the responsibility of not only becoming AI-fluent but also learning how to apply this knowledge across diverse fields, understanding that the future competition is not human versus machine, but an AI-augmented colleague versus another.