- Workers between 22 and 25 years old seem to be the most affected by AI
- Software development and customer service are strongly affected
- Roles subject to higher automation are the most at risk
A new article by researchers from the University of Stanford discovered six truths on the effects of AI on the job market, and it may not be so good for young workers.
Data show that young workers, aged 22 to 25, in the jobs most exposed to AI have experienced a substantial drop in employment since the end of 2022 – this includes roles such as software development and customer support.
According to the researchers, the researchers have led to a stagnant job overall despite the total American employment rates that continue to increase.
AI could take younger jobs
In July 2025, for example, employment for software developers in this age group fell by almost 20% compared to the end of 2022. Overall, employment in the roles most exposed to AI decreased by around 6% for this young demography, but older workers (defined as 35-49) saw an increase of 6-9%.
Stanford’s newspaper somehow explains why youth employment was relatively stable despite global national growth.
Brynjolfsson, Chandar and Chen – The researchers behind the document – goes further by dividing AI into two distinct categories – automation and increase.
Young workers were the most affected by AI as an authorization, which replaces tasks and leads to a drop in entry -level jobs.
On the other hand, elderly workers were more likely to be affected by AI as an increase, where it supports human work. In this case, researchers have not seen any decline, and sometimes even growth.
They found that employment rates were harder than wages in most cases, the reductions in the workforce more likely to occur than salary reductions. Already this calendar year, the technological industry has experienced more than 81,000 layoffs, although this is down compared to a summit of 264,000+ of 2023 (for the whole year).
However, the document suggests that all hope may not be lost, pointing to previous trends such as the IT revolution which “finally led to a solid growth in employment and real wages after physical adjustments and human capital”.
In this spirit, it is possible that AI can indeed improve the labor market market, but only after a first period of turbulence which affects workers less qualified in a disproportionate manner.
