Junior employment falls by 7.7% in companies adopting AI in the six quarters following implementation. This silent erosion structurally transforms professional trajectories without mass layoffs, closing the entry doors of the labor market to an entire generation.
The essentials
- Junior employment declines by 7.7% in the 18 months following AI adoption, according to a Harvard study of 62 million American workers
- Companies stop hiring beginners rather than laying off, creating an invisible transformation
- Senior employment continues to grow by 3.2% in the same companies, intensifying the generational divide
- This phenomenon particularly affects administrative, accounting, and support professions
Companies quietly shut off the junior tap
Artificial intelligence does not cause the predicted catastrophe. It operates differently: by cutting off hiring of beginners. The Harvard Business School study conducted by Seyed M. Hosseini and Guy Lichtinger on 62 million American workers reveals an invisible but radical mechanism.
Companies deploying generative AI reduce their hiring of junior profiles by 7.7% in the 18 months following implementation. At the same time, employment of experienced workers increases by 3.2%. This strategy avoids the social and political costs of layoffs while fundamentally restructuring the age pyramid.
The transformation accelerates as AI grows more sophisticated. Positions for assistants, junior analysts, and coordinators disappear from job postings. Companies prefer to entrust these tasks to AI and concentrate their human teams on senior expertise.
The data reveal a contrasting geography of this mutation. Technology, finance, and consulting sectors experience the strongest contractions in junior hiring, with declines reaching 12% in certain firms. Conversely, sectors requiring physical presence—healthcare, education, and crafts—maintain their entry-level recruitment.
Social mobility grinds to a halt at the bottom
This closure of the junior market threatens traditional mechanisms of social ascension. Entry-level positions served as practical training schools, enabling graduates from all backgrounds to acquire the experience valued by employers. Their disappearance creates a paradox: how can one acquire the required experience if nobody is hiring beginners anymore?
Companies now concentrate their hiring on profiles with a minimum of 5 to 10 years of experience. This strategy exacerbates the shortage of intermediate talent while closing the door to newcomers. The labor market polarizes between an experienced elite increasingly courted, and a mass of graduates without entry opportunities.
The impact strikes particularly hard at graduates from modest backgrounds, who cannot afford prolonged unpaid internships or costly additional training. These individuals find themselves excluded from a market that demands experience they cannot acquire. The meritocracy promised by higher education crumbles before this new technological barrier.
European data confirm this trend. The European employment observatory notes a 15% drop in permanent contract hiring for those under 25 in tech companies between 2023 and 2025. France, Germany, and the Netherlands record similar developments, suggesting a structural phenomenon transcending national specificities.
Experience becomes the new scarcity
Paradoxically, this rarefaction of junior hiring transforms experience into a precious asset. Professionals who survived the first waves of automation see their value skyrocket. Salaries for experienced profiles rise by 8 to 15% in sectors massively adopting AI, according to compensation data from recruitment firms.
This inflation of senior salaries creates a spiral: the more expensive experience becomes, the more companies invest in AI to reduce their dependence on human skills. The more AI develops, the fewer juniors they hire. The circle closes on a sacrificed generation.
Companies develop new internal training strategies. Rather than hiring beginners, they recycle their experienced employees toward new professions. These retraining programs primarily concern those over 35, leaving young graduates out of this reshuffling of professional roles.
This dynamic also transforms intergenerational relationships at work. Seniors, once threatened by their skills becoming obsolete, recover the status of an indispensable mentor. Their experience of the “pre-AI era” becomes critical expertise for guiding technological integration. This inversion of power dynamics questions companies’ generational diversity policies.
Interaction-based professions resist better
Not all junior jobs suffer the same erosion. The Harvard study reveals significant disparities depending on the nature of positions. Professions requiring complex human interactions—sales, negotiation, project management—maintain junior hiring rates close to pre-AI levels.
This resistance is explained by AI’s difficulty in reproducing relational intelligence. A junior salesperson brings empathy and adaptability that algorithms struggle to match. Similarly, creative positions—design, marketing, communication—maintain their entry-level hiring, with AI serving as an assistant rather than a substitute.
Technical professions experience a more contrasted fate. Software development sees its junior hiring fall by 18%, with generative AI automating simple code production. In contrast, hardware engineering and energy transition professions recruit large numbers of young graduates, demand far exceeding the supply of experienced profiles.
This geography of opportunities redraws career guidance. Engineering schools are reorienting their curricula toward specialties resistant to automation: robotics, renewable energy, biotechnology. Training in humanities and social sciences regain unexpected appeal, their graduates mastering the relational skills valued by companies.
Public policies adapt slowly
Facing this silent mutation, governments struggle to adjust their employment policies. Traditional indicators—unemployment rate, net job creation—mask the reality of frozen junior hiring. Youth employment support policies, designed for an economy where companies lay off then rehire, prove inadequate to this new situation.
Some countries are experimenting with innovative responses. Denmark is testing a “right to first experience” requiring companies with more than 500 employees to dedicate 5% of their hiring to profiles without significant experience. The Netherlands is developing an enhanced tax credit system for hiring juniors in sectors undergoing digital transformation.
France is launching a pilot program in 2026 for long-term paid internships in administrations, allowing young graduates to acquire the experience demanded by the private sector. These mechanisms remain marginal compared to the scale of the challenge. The transformation of the labor market requires a complete overhaul of training-to-employment transition mechanisms, still in embryonic form in most developed countries.
The challenge transcends national borders. The European Union is considering a common regulatory framework for AI in companies, including obligations to maintain a minimum threshold of junior hiring. These negotiations promise to be complex, with companies threatening to relocate their activities to less restrictive jurisdictions.
A generation sacrificed on the altar of efficiency
The great hiring freeze creates a new generational fracture. On one side, a generation of people in their forties and fifties valued for their pre-AI experience. On the other, qualified young graduates excluded from a market that no longer wants to train them.
This evolution questions the social contract around work. The idea that educational effort guarantees access to employment erodes before technological barriers that individuals do not control. Extreme poverty ceases to decline for the first time in thirty-five years, creating a context where this exclusion of young people from the labor market takes on a dramatic social dimension.
The consequences go beyond the economic. A generation deprived of professional experience develops different life trajectories: delays in residential independence, postponement of family projects, prolonged dependence on parents. These societal transformations accompany the technological revolution without their implications being fully measured.
Artificial intelligence keeps its promise of economic efficiency. But this efficiency has a social cost: the closure of professional elevators for an entire generation. Between algorithmic optimization and intergenerational justice, developed societies navigate an arbitration whose outcome will determine the social cohesion of the decades to come.