The rise of artificial intelligences like ChatGPT, launched in late 2022, promises to boost productivity, but it also raises legitimate concerns about the future of careers and employment.
Research led by economists at Stanford University highlights a phenomenon: young professionals aged 22 to 25 are paying the highest price for this transformation. Their entry into the job market is stagnating—or even declining—in the sectors most vulnerable to AI, while their older peers continue to move forward. This imbalance is not trivial; it calls into question the very foundations of social mobility and vocational training.
This wave of automation is replacing routine tasks often assigned to beginners, limiting their first steps in the workplace.
AI and youth employment: key findings from the Stanford University study
Based on ADP administrative payroll data (the largest provider in the United States), Stanford’s analysis reveals six main findings:
- Relative 13% drop: Young people in the occupations most exposed to AI are seeing a 13% decline in employment.
- Stagnation since 2022: Overall, their employment trend has stalled, with an absolute 6% contraction in these sectors.
- Growth for older workers: More experienced workers in the same jobs see employment rise by 6% to 9%.
- Adjustments via headcount: Companies cut positions rather than wages, highlighting wage rigidity.
- Automation vs. augmentation: The decline is concentrated in roles where AI replaces human work, not where it assists it—such as codified tasks that are easy to automate.
- Robustness: The results hold after excluding tech firms and remote-workable jobs, confirming a direct link with AI adoption.
In protected sectors, such as healthcare or maintenance, young people follow an upward trajectory similar to their older peers, showing the impact is not uniform.
Also read on this topic: TOP 10 in-demand jobs in artificial intelligence and their salaries in 2025
Impacts and examples
These young talents are seeing their opportunities shrink: for junior developers, employment fell by nearly 20% between late 2022 and mid-2025, while chatbots now handle basic customer service interactions. This generational gap could slow the development of future experts, since entry-level roles traditionally serve as a springboard.
As Erik Brynjolfsson, co-author of the study, explains, young people need to learn to work with AI to boost their productivity, rather than be subjected to it.
These early signals point to a disproportionate impact on newcomers, potentially slowing long-term innovation and calling for tailored training starting at university.
Source: digitaleconomy.stanford
