Fewer job opportunities for young people competing with AI
Generative AI is starting to affect the Dutch labor market, Rabobank economists wrote in the economic journal ESB. There are fewer vacancies and employment opportunities in professions susceptible to automation by AI, and this is primarily affecting young people. Employment in these professions decreased by 13 percent for young workers under 25 since the advent of artificial intelligence.
The Rabobank researchers took the 114 professions classified by Statistics Netherlands (CBS) and assessed them for how susceptible they would be to being automated by Generative AI. They found that professions with many higher professional or university graduates are more susceptible to such automation.
They then compared vacancy and employment data from CBS and benefits agency UWV from before the widespread use of AI (late 2022) and the most current available data (second and third quarter 2025).
The number of vacancies in the occupational groups most suitable for GenAI automation has lagged behind the rest of the labor market in recent years, the economists found. Until 2023, the number of new and open vacancies in these professions kept pace with other occupational groups. But since the advent of AI, both the number of new and open vacancies has declined “significantly” faster than with other occupational groups.
The number of open vacancies in the most susceptible professions was almost 25 percent lower in the second quarter of 2025 than at the end of 2022. In the other 102 occupational groups, the number of vacancies decreased by less than 10 percent in that period. “All in all, this represents a decrease of over 19,000 vacancies in GenAI-eligible occupational groups,” the researchers said.
CBS employment figures show that employment for young people aged 15 to 24 has declined sharply in GenAI-eligible occupational groups. In the third quarter of 2025, over 13 percent fewer young people were working in these groups than in the fourth quarter of 2022. In the other occupational groups, youth employment rose by 3 percent. The gap is much smaller for other age groups.
“Although employment and unemployment figures are not directly comparable, it is conceivable that the rise in GenAI within Dutch companies contributed to the recent resurgence of (youth) unemployment in the Netherlands,” the researchers said. Approximately 19,000 fewer 15- to 24-year-olds were employed in GenAI-sensitive jobs in the third quarter of 2025 than in the last quarter of 2022. In that same period, the number of unemployed Dutch people increased by 38,000, including 25,000 young workers.
According to researchers, other trends do not adequately explain the recent decline in youth employment in GenAI-sensitive occupations. The fact that young people in particular seem to be affected has to do with how AI works. GenAI is trained with codified knowledge, knowledge learned from books and manuals, and has no tacit knowledge, knowledge gained from practical insights through experience. “This means that, in theory, GenAI competes primarily with young workers, who, fresh out of school, primarily possess codified knowledge, but little practical experience.”
The researchers stressed that this does not mean that GenAI will have a lasting negative effect on employment. “It is conceivable that other sectors facing labor shortages will benefit from the increased labor supply,” the economists wrote. “Furthermore, it is possible that, if GenAI does indeed increase productivity, it will make certain products or services less labor-intensive and therefore cheaper. This can boost demand for these products/services, which supports employment. Moreover, new technologies often create new jobs, which are currently unknown.”
