• AI adoption is rapidly reducing demand for entry-level roles in high-exposure occupations, with early-career workers showing measurable declines in employment.
• Payroll and job postings data indicate that firms are skipping junior hires and concentrating demand in mid-career talent.
• This structural shift creates investment opportunities in sectors benefiting from leaner workforce models while posing risks to consumer-dependent industries.
Over the last two years, artificial intelligence (AI) adoption has shifted from theoretical to practical, with direct consequences for the labor market. Data now indicates that early-career workers in occupations most exposed to large language models are experiencing measurable employment declines. This is not a broad-based cyclical slowdown, but rather a structural change in the way firms hire and allocate junior talent.
Evidence from Payrolls
A recent study using ADP payroll data covering 50 million workers finds that early-career employees (ages 22-25) in occupations with high AI exposure experienced a 13% relative decline in employment between November 2022 and July 2025.1 Importantly, this effect persists after accounting for firm-level and industry shocks, indicating it cannot be fully explained by higher interest rates or macroeconomic weakness.
Job Postings and the Shift in Experience Mix
The decline in payroll employment is corroborated by job postings data. The Burning Glass Institute reports that entry-level postings (less than 3 years of experience) in high-AI-exposure occupations fell from 2022 to 2025, while postings requiring 6+ years of experience held steady or rose.2 This suggests employers are not reducing demand for the occupation itself, but are instead eliminating the junior tier, as many of the primary job functions of entry-level employees are already well within AI’s capabilities.
Employer Plans Cooling Quickly
In the fall of 2024, employers reported plans to hire 7.3% more graduates from the Class of 2025 than they did from the Class of 2024. As of spring 2025, that expectation had been cut to just 0.6%.3 This steep downshift suggests employers are rethinking near-term graduate hiring needs, even as overall employment levels remain steady. Combined with evidence that the sharpest pullback is in AI-exposed occupations, this reinforces the view that structural forces, rather than broad macro weakness, are shaping demand for entry-level talent.
The Long-Term Risk
The St. Louis Federal Reserve has summarized research showing that approximately 52% of graduates are underemployed in their first role, and that this condition is predictive of weaker wage outcomes for a decade thereafter.4 With AI compressing entry-level opportunities in high-exposure occupations, more graduates are likely to be diverted into roles outside their field of study, raising the risk of persistent wage scarring.
Technological disruptions have historically created new job categories even as they eliminated others. The rise of the internet destroyed travel agencies but created entire industries around e-commerce and digital marketing. Some economists argue that AI will follow a similar pattern, eventually generating roles we cannot yet envision. However, unlike previous technological shifts that played out over decades, AI’s impact on cognitive work appears to be happening much faster, potentially outpacing the economy’s ability to create offsetting opportunities for new graduates.
Investment Implications
Professional services firms, investment banks and technology companies that have historically relied on large analyst classes may move to “leaner pyramids,” improving margins significantly. Consulting firms like McKinsey and technology companies with high junior-to-senior ratios could see meaningful cost reductions. Financial services firms that can maintain deal flow and client coverage with fewer entry-level employees stand to benefit from improved operating leverage.2
On the other side of the ledger, lower incomes among younger cohorts may weigh on housing formation and discretionary consumption. Sectors particularly exposed to young adult spending, including entry-level housing, furniture and certain retail categories, face potential demand compression. Student housing and educational services may also see pressure as the return on investment for higher education diminishes.
Mid-career workers with the skills to direct and validate AI outputs may see a wage premium expand, while new entrants face fewer opportunities. This suggests potential outperformance for companies with strong training programs and employee retention, as they’ll be better positioned to capitalize on scarce mid-level talent.
Market Sectors at Risk
Higher education institutions face particular pressure. If enrollment levels decline due to weakened employment prospects, universities with high dependence on tuition revenue could face financial stress. Conversely, institutions that successfully pivot toward experiential learning and “AI-plus” curricula may gain market share by presenting employers with candidates who can contribute at higher levels earlier in their careers.
The student lending market also faces headwinds. With weaker employment outcomes for graduates, loan repayment rates could deteriorate, affecting both government programs and private lenders. This risk is compounded by the fact that underemployment tends to persist, making it a structural rather than cyclical concern.
Looking Ahead
If entry-level opportunities remain limited, the next generation of workers will adapt by pursuing alternative entry ramps such as short-term contract work, entrepreneurial ventures or AI-assisted freelancing that builds experience outside of traditional firms. While this may create a more flexible workforce, it also risks widening disparities between graduates who can navigate fragmented career paths and those who remain underemployed.
For firms, eliminating the junior tier could produce a talent bottleneck later in the decade. Employers currently seeking candidates with 5-10 years of experience may find that such workers are in short supply, forcing higher wages or renewed investment in structured training pipelines.
The near-term evidence shows that AI is already reshaping the structure of hiring, reducing demand for entry-level positions in high-exposure occupations. For markets, the aggregate employment effect may remain muted, but the distributional consequences—on consumption patterns, sector margins and labor supply imbalances—are likely to intensify as this structural shift continues.
Important Disclosures & Definitions
1 Brynjolfsson, E., Chandar, B., & Chen, D. (2025). Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of AI. Stanford Digital Economy Lab Working Paper.
2 Levanon, G., Sigelman, M., Mamertino, M., de Zeeuw, M., & Guilford, G. (2025). No Country for Young Grads: The AI Disruption of Entry-Level Jobs. Published by Burning Glass Institute.
3 National Association of Colleges and Employers. (2025). Job Outlook 2025: Spring Update.
4 Federal Reserve Bank of St. Louis. (2025). Underemployment and College Graduates. Open Vault publication.
AAI000987 09/02/2026
