Generative AI’s Profound Early Impact on Entry-Level Jobs: Stanford Study Reveals 13% Decline Among Young Workers

In a landmark study released by Stanford University’s Digital Economy Lab, researchers Erik Brynjolfsson, Bharat Chandar, and Ruyu Chen have uncovered compelling evidence that generative AI is already reshaping the American labor market—starting with its youngest members. The paper, titled “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence,” uses high-frequency payroll data from ADP, the largest payroll software provider in the U.S., to track employment trends across millions of workers. The findings are stark: entry-level workers aged 22–25 in AI-exposed occupations have experienced a 13% relative decline in employment since late 2022.

This decline, the researchers argue, is not a temporary blip or a result of broader economic conditions. Instead, it marks the beginning of a structural shift driven by the rapid adoption of generative AI tools—particularly in roles where AI is more likely to automate rather than augment human labor.

Six Key Findings That Signal a New Era

The study presents six core facts that collectively paint a picture of generative AI’s early labor market effects:

1. Sharp Decline for Young Workers in AI-Exposed Roles Occupations such as software development and customer service—both highly exposed to generative AI—have seen significant employment drops among workers aged 22–25. In contrast, older workers in the same roles and younger workers in less-exposed fields (e.g., nursing aides) have maintained or grown their employment levels.
2. Overall Employment Growth Masks Youth Stagnation While the broader labor market continues to grow, employment for young workers has stagnated. In AI-exposed jobs, young workers saw a 6% decline in employment from late 2022 to mid-2025, while older cohorts experienced a 6–9% increase.
3. Automation Drives Decline—Not Augmentation The study distinguishes between AI applications that automate tasks versus those that augment human capabilities. Entry-level employment has declined in roles where AI substitutes for labor, but has grown in occupations where AI complements human work.
4. Firm-Level Shocks Don’t Explain the Trend Even after controlling for firm-specific economic shocks, the employment decline among young workers in AI-exposed roles remains statistically significant. This suggests that the trend is directly tied to AI’s influence rather than broader macroeconomic factors.
5. Compensation Remains Stable—For Now Interestingly, while employment has dropped, wages have not. The researchers found little difference in salary trends across age groups or exposure levels, indicating that AI’s impact is currently more visible in hiring than in pay.
6. Robust Across Industries and Job Types The findings hold even when excluding tech-related firms and remote-friendly occupations. This suggests that AI’s disruptive effects are not confined to Silicon Valley or telework roles, but are spreading across diverse sectors.

Why Young Workers Are More Vulnerable

The authors offer a compelling hypothesis for why entry-level workers are bearing the brunt of AI’s disruption. Generative AI systems are trained on codified knowledge—the kind typically acquired through formal education. Young professionals, fresh out of college or vocational training, tend to rely heavily on this codified skill set. In contrast, older workers possess more tacit knowledge—practical, experience-based insights that are harder for AI to replicate.

This dynamic means that AI is more likely to replace the tasks performed by younger workers, while older employees may be shielded by their accumulated expertise. The study also notes that in occupations with low returns to experience, even non-college workers up to age 40 are beginning to feel the effects.

Implications for Policy, Education, and Workforce Strategy

The Stanford study arrives at a critical moment. With nearly half of U.S. adults reporting some use of generative AI at work by mid-2025, and tools like ChatGPT and Claude becoming ubiquitous, the pressure to adapt is mounting.

For policymakers, the findings underscore the need to rethink workforce development strategies. Traditional education may no longer be sufficient to prepare young workers for a labor market increasingly shaped by AI. Instead, there may be a growing need for experiential learning, mentorship, and upskilling programs that emphasize tacit knowledge and human-AI collaboration.

Employers, too, face a strategic crossroads. While AI can boost productivity, it may also erode the pipeline of early-career talent. Companies that rely heavily on automation risk losing the long-term benefits of nurturing junior employees who grow into seasoned professionals.

A Call for Ongoing Monitoring and Research

The authors caution that their findings, while robust, represent an early snapshot of a rapidly evolving landscape. They plan to continue tracking employment trends to determine whether these patterns accelerate, stabilize, or reverse over time. Future research will benefit from more granular data on firm-level AI adoption and task-level disruption. This reminds me of another study done by Microsoft on the Early Effects of M365 Copilot that was released back in April.

In the meantime, the message is clear: generative AI is no longer a distant threat or abstract concept. It is already reshaping the entry-level job market—and young workers are the first to feel the tremors.

Download the Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence PDF

Related Posts

1. Microsoft Brings Convenient Microsoft 365 Companions to Windows 11, Now in Beta
2. Microsoft Education Rolls Out Major Exciting Back-to-School Updates: Copilot AI Features, LMS Integration, Learning Accelerators, and More for August 2025
3. Microsoft Copilot Gets Major Upgrade in August 2025: GPT-5 Integration, WhatsApp Agent Deployment, and Advanced Enterprise-Grade Security
4. Microsoft Releases August 2025 Exchange Server Security Updates: What IT Pros Need to Know
5. 2025 OpenAI Restructuring in Limbo: Microsoft Negotiations Stall, IPO and Bigger Funding at Risk