Unemployment among IT graduates from universities of applied sciences has more than doubled in two years in Switzerland. Job vacancies for junior positions, meanwhile, fell by 32% in 2025, with IT professionals being the most affected.
The cause is structural, not cyclical: the basic tasks that young developers traditionally performed to learn the profession are now being assigned to generative AI. Companies now expect graduates to have much stronger strategic and interpersonal skills than before, rather than merely a basic technical command.
For recruiters, this means revisiting junior hiring criteria: a coding test is no longer enough to assess a candidate. For applicants and schools, a warning sign is also emerging in terms of learning: researchers refer to “learning penalties,” meaning slower learning among young people who rely too early and too systematically on AI to solve their first technical problems.
This week: if you are an IT student or recent graduate, take an exercise or bug you solved recently and redo it entirely without AI, from diagnosis to solution. Prepare to explain it in an interview: today, it is the ability to reason independently—not the speed of code generation—that makes the difference.
LinkedIn post idea to adapt
“Unemployment among young IT graduates has doubled in two years in Switzerland, while junior vacancies have fallen by 32%. This is not the end of the profession; it is the end of one way of entering it. Recruiters: ask your junior candidates to solve a problem in front of you, without AI. You will quickly see who has learned to think and who has learned to prompt.”