Ai is erasing the junior rung of the corporate ladder
The entry-level job is quietly vanishing. A BBVA study tracking 14 million LinkedIn profiles shows that companies now hire 28 % fewer graduates for roles that generative AI can already handle—code reviews, legal discovery, financial modelling—while senior head-count keeps climbing. The pattern is global, spans banking to biotech, and demolishes the apprenticeship model that fuelled post-war career mobility.
Lo que nadie cuenta es que the contraction is not cyclical; it is structural. BBVA’s chief economist Rafael Doménech filters out recession noise and still finds a 13 % relative drop in employment for workers under 28 in AI-exposed occupations since ChatGPT’s release. Promotions have not accelerated to fill the gap. Firms simply stopped creating junior seats.
The canary data: spain, canada, uk
Spain’s youth unemployment rate hovers at 26 %, Eurostat’s worst. In Canada the share of 20- to 24-year-olds employed has slid to its lowest since 1998. UK job board Adzuna reports a 32 % plunge in graduate-level adverts since November 2022. Meanwhile vacancies for senior architects, drug-development leads and prompt-engineering specialists are up 9 % year-on-year across the three markets.
The shift is starkest in software. U.S. Bureau of Labor Statistics microdata analysed by Stanford’s Erik Brynjolfsson show a 20 % crash in head-count for developers aged 22-25 between 2022 and 2024 even as total tech payrolls expanded 4 %. Anthropic CEO Dario Amodei forecasts half of all entry-level white-collar roles gone by 2030, translating into a 10-20 % unemployment spike for degree holders.
Why seniors survive
Large-language models ingest code, contracts and spreadsheets but choke on organisational politics, client nuance and the tacit knowledge acquired by debugging a legacy trading floor at 3 a.m. Senior staff act as guardrails, not targets. Their wage premium is widening: average compensation for 40-year-old quant developers at Goldman Sachs rose 18 % last year while intern stipends froze.
Companies pocket a 4 % productivity dividend from AI, BBVA calculates, yet channel the gains into shareholder buybacks and senior talent rather than trainee pipelines. The calculus is brutal: a junior analyst costs $85 k fully loaded and needs two years to break even; a fine-tuned LLP costs $3 k per month and never asks for paternity leave.
The decade-long scar
Missing the first rung leaves a mark. A Center for American Progress model shows six months of joblessness at 22 cuts lifetime earnings by $22 000 in real terms. Multiply that by an entire cohort and the inter-generational wealth gap yawns wider. Spain already loses 25 % of its graduates to emigration within three years; AI is set to accelerate the brain drain.
Policy tinkering—tax credits for hiring apprentices, subsidies for reskilling—collides with corporate incentives that reward quarterly margins over five-year talent pipelines. Universities keep selling tuition on the promise of a first job that, in many desks, no longer exists.
The final insult: the sectors immune to algorithmic substitution—construction, hospitality, nutrition—pay median wages 30 % below those of the disappearing junior analyst track. graduates can find work, just not the kind that repays their loans or keeps them in the cities where power and capital concentrate.
Wall Street’s summer internship programmes open next month with 40 % fewer slots. The résumés will still pile up, but the algorithm has already read them—and written the rejection before HR wakes up.