AI is not firing young workers, it is just not hiring them
Stanford researchers found a 19 percent employment gap for young workers in AI-exposed jobs because companies stopped hiring beginners.
Hiring · Source: Stanford Digital Economy Lab
What happened
Stanford researchers updated their payroll study on how AI affects jobs. They found no widespread job losses across the economy. However, young workers in roles exposed to AI are falling behind. The gap exists because companies are hiring fewer beginners, not because they are firing them.
The data shows a 19 percent employment gap for workers aged 22 to 25 in highly AI-exposed jobs as of June 2026. This gap has widened steadily since the summer of 2025. The drop is concentrated in roles where AI automates routine tasks based on formal, documented knowledge.
Experienced workers do not face this same gap. In jobs that rely on tacit knowledge gained through practice and mentorship, employment is actually flat or rising. The New York Fed reports that unemployment for recent college graduates sits at 5.6 percent, with underemployment at 42 percent.
By the numbers
- 19% — Employment gap for 22-to-25-year-olds in highly AI-exposed jobs as of June 2026
- 5.6% — Unemployment rate for recent college graduates aged 22 to 27 in the second quarter of 2026
- 42% — Underemployment rate for recent college graduates in the second quarter of 2026
- $5,000 — Penalty given to lawyers who filed a brief citing fake court opinions invented by ChatGPT
Why it matters
The traditional entry-level job is disappearing in fields where AI can do the basic work. Companies are relying on AI to draft reports, write code, and answer customer questions. This means new graduates can no longer rely on doing tedious, repetitive tasks to get their foot in the door and learn the business.
The new entry-level job is about checking the work that AI produces. When an AI makes a mistake, the company is still held responsible, just like Air Canada and the lawyers in the Avianca case. Junior workers must now build their judgment by verifying AI outputs against original sources and explaining the results to their managers.
Who this hits
New graduates
This group faces a 19 percent employment shortfall in jobs highly exposed to AI. They must prove they can manage and check AI outputs rather than just perform basic tasks.
Experienced workers
Older workers with tacit knowledge are not seeing the same hiring drop. Their employment is flat or rising because their experience-based judgment is harder for AI to replicate.
Women
Stanford researchers note that women face greater AI exposure on average in the workforce. This is a trend the researchers plan to monitor closely as the data evolves.
What to do this week
- Stop applying only to traditional entry-level postings and look for roles that value hands-on experience and tacit knowledge.
- Practice using AI tools to generate drafts, then manually check every fact and number against the original source.
- Learn to explain your work and the AI output step by step so a manager can trust your judgment.
See who's hiring right now on GiraffyReach
My take
I see many new graduates worrying about AI taking their jobs, but the door is not slamming shut. It is just changing shape. If you want to get hired today, you need to show employers that you can catch the mistakes the AI makes and take full responsibility for the final result.