Starting out in the AI era: the entry-level squeeze and how to get around it.

The squeeze

Stanford's Digital Economy Lab tracks millions of US payroll records. By June 2026, employment of 22 to 25 year olds in the most AI-exposed occupations was 19% below the level their trend in less exposed work would predict, widening from 15% a year earlier. In level terms the two most exposed groups fell about 11% since late 2022 while the three least exposed rose about 10%. ADP's own read is blunter: the youngest software developers were 20% below their late 2022 peak, early-career customer service workers nearly 11% below.

Two details matter more than the headline. First, the drop comes from reduced hiring, not from people being let go. Second, it is concentrated where AI automates the work. Where AI augments the work, employment of young people is flat or rising.

The counterweight

The most cited field study of generative AI at work, across more than 5,000 customer support agents, found a 14% average productivity gain and a 34% gain for the least experienced workers. AI compresses the gap between a new starter and a good one, because it carries the accumulated know-how of the good ones. If you are early in your career and using AI properly, you are the person the study is describing.

How to get around it

Aim at augmented roles, not automated ones. If the job listing reads like a task list, the task list is what gets automated. If it reads like a set of responsibilities to people, it is safer. Show the compressed learning curve in the interview: bring the thing you built, and explain where the tool was wrong and how you caught it. Employers are worried about exactly that judgement, and most candidates cannot demonstrate it.

Common questions

Is AI reducing entry-level jobs?

In the most AI-exposed occupations, yes. Stanford found employment of 22 to 25 year olds 19% below trend by June 2026, driven by fewer hires rather than more separations. Less exposed occupations grew about 10%.

Do new workers benefit from using AI?

Yes. The largest field study found a 34% productivity gain for the least experienced workers, against 14% on average (Brynjolfsson, Li and Raymond).

How do I make my application stand out in the AI era?

Aim at augmented roles, show work you built with AI and explain where the tool was wrong, and carry a measured readiness score rather than a claim.