AI adoption erodes entry-level pathways as young workers face structural labour market shift
With graduate unemployment at 5.6% and underemployment soaring to 42.5% in late 2025, experts warn that automating junior tasks threatens long-term workforce capability and institutional memory.
A structural shift is reshaping the labour market as artificial intelligence increasingly substitutes for entry-level tasks, potentially hindering career progression for young workers. A working paper from the Stanford Digital Economy Lab, released in November 2025, reports a 16% relative decline in employment for workers aged 22 to 25 in AI-exposed occupations. This trend coincides with a softening graduate labour market, where the Federal Reserve Bank of New York recorded a 5.6% unemployment rate and 42.5% underemployment for recent college graduates in the fourth quarter of 2025.
The decline is specific to early-career jobs with high AI exposure, such as software development, customer service, programming, and information systems management. An Anthropic report from March 2026 provides suggestive evidence corroborating these findings, noting that more experienced workers in the same occupations did not suffer similar employment drops. While aggregate employment in developed countries remains broadly stable, experts warn that the traditional pathway into these professions is becoming less viable as AI handles routine coding and administrative tasks.
Georgios Petropoulos, an assistant professor at the USC Marshall School of Business, links this shift to broader concerns about long-term workforce capability and institutional memory. He argues that traditional entry-level roles serve as a critical training system for junior analysts, developers, and legal or financial staff to learn practical judgment, system failures, and human relationship dynamics. If AI absorbs the drafting, triage, and coding that once helped train these workers, firms may gain short-term efficiency at the cost of long-term organisational knowledge.
Authorities are urging educational institutions, governments, and businesses to adapt by embedding AI literacy into degrees and incentivising early-career hiring. The focus is shifting from routine coding to supervising AI systems, with domain expertise combined with AI fluency emerging as the new scarce commodity. Experts suggest that universities should require verification skills and domain judgment in ordinary degrees, while governments could utilise existing tax policy architectures to create subsidies for employers hiring early-career workers into structured, AI-augmented roles.
Students entering the workforce now face a tough transition where AI fluency is becoming a commodity, but the combination of domain expertise and AI proficiency remains highly valued. The competition for young workers is increasingly between a colleague and an AI-augmented colleague, requiring graduates to understand how to apply AI tools within specific fields rather than relying on technical skills alone. Firms are being urged to view entry-level hiring not just as a cost-saving measure but as an essential investment in the future stock of judgment and productivity within their organisations.
