Research opportunity · Aggregate null or distributional shift? Employment
Are juniors who still get hired at AI-adopting firms promoted and trained less?
Earlier version (v1)
This is RO-LD02-04 as it was published at v1, kept unchanged. The current version is v2; its changes are listed there.
Question
Among workers hired into junior positions before and after their employer began using AI, do promotion into higher-skill occupations, wage growth, recorded training and retention over the following two to four years differ between AI-adopting and comparable non-adopting firms?
Causal mechanism not isolatedOne PhD chapter or paperLow to medium costPartly addressed
Why it matters
The worry behind "eroding career ladders" is not only that fewer juniors are hired, but that those who are hired learn less, because AI now does the routine work through which novices used to learn. Current studies mostly count junior hires and junior shares; the one that looks at junior promotions covers only a short window, and none found follows the juniors inside adopting firms over several years. If promotion and wage growth hold up, the ladder is narrower but intact; if they fall, the pipeline to senior roles is at risk.
The debate lists this among what we still do not know: Will entry-level learning ladders erode if firms hire fewer juniors?
Evidence that motivates it
- In a 41-country working paper, AI-adopting multinational affiliates shift toward more senior workforces: junior share falls by 1.9 percentage points, driven mainly by 6.7% senior employment growth rather than a statistically clear decline in junior headcount.
In a 41-country working paper, AI-adopting affiliates shift toward more senior workforces mainly through senior growth, not a clear fall in junior headcount; the claim does not license junior job destruction, so the open question is the quality of junior careers, not their number. - Danish firms that began using AI in 2023 had roughly 11% lower relative employment growth by late 2025 than comparable non-adopters, mainly through lower recruitment, while economy-wide employment remained relatively unaffected.
Danish firms that began using AI in 2023 grew employment about 11% less than non-adopters, mainly through lower recruitment; this identifies a register population of adopting firms whose junior hires can be followed, and it does not license reading the gap as layoffs.
A possible design
Matched employer–employee panel. Identify firms that began using AI from an official enterprise ICT-usage survey linked to registers, and match them to non-adopters on size, sector, pre-period growth and occupational mix. Take every worker hired into a junior position (by age, experience or occupational level) in 2018–2025. Compare, by hiring year and adopter status, the probability of moving to a higher-skill occupation code within the firm, wage growth, separation, and, where the register records it, participation in employer-related training. Use a difference-in-differences across pre- and post-adoption hiring cohorts. A small set of interviews with managers in adopting firms can show how junior tasks were reallocated.
Data needed
Linked employer–employee registers with occupation codes and earnings (Denmark, Norway, Sweden, Finland or the Netherlands); firm-level AI-use answers from the national ICT-usage-in-enterprises survey linked by firm identifier; training records where they exist (coverage varies by country and must be checked).
Feasibility
One PhD chapter or paper. Low to medium; register access fees and time for approvals; interviews add little cost. The harmonised EU version of the ICT survey covers only enterprises with at least 10 employees and is a sample, and its AI questions differ across years, so adoption timing may be coarse and should be checked wave by wave for the chosen country. Occupation codes change slowly within firms, so promotion must be measured as code changes plus wage jumps, and the measure should be validated before use.
Has this been done?
Partly done already: related work covers some of this question (first-pass screen of 12 October 2026; not yet checked in depth).
US résumé data show that adopting firms hire fewer juniors with no significant short-window change in junior promotions, and the authors list longer-run adjustments in training and career ladders as an open question. The 41-country paper measures affiliate headcounts and junior shares, not individual trajectories, and the Danish central-bank paper reports firm employment growth and recruitment, according to the pages read. No study was found that follows juniors hired at adopting firms over several years for promotion, wage growth and training in population registers.
How strong this evidence is
- Who searched and checked: an independent AI reviewer with fresh context (not a person), 12 October 2026: every nearest-work citation checked against its source page; corrections applied. Every check so far was done by AI; no person or domain expert has checked it yet.
- What was read: 1 of the 4 nearest works below was judged from an abstract, summary or search listing only, and 1 has details that could not be confirmed; it is marked. Not every entry says how much of the work was read.
- Where the search could not reach: Screened by web search on the date shown; Google Scholar, SSRN and paywalled journal pages were not systematically searched, and the Danish central-bank paper was read at landing-page level only. Several small interview studies of junior software developers exist and were not treated as nearest work because they do not measure careers. Absence of a hit is not proof that the work does not exist.
- Before building a project on it, run your own search in a scholarly database such as Scopus, Web of Science, Google Scholar or SSRN.
Nearest work found
- Hosseini Maasoum & Lichtinger, "Generative AI as Seniority-Biased Technological Change: Evidence from U.S. Résumé and Job Posting Data," SSRN Working Paper No. 5425555 (2025)
Revelio résumé data, adoption flagged by GenAI-integrator postings: relative to non-adopters, adopters hire about 5 fewer juniors per quarter after early 2023, and junior promotions show no significant change in that short window. Does not estimate effects on training, skill development or wage growth. Figures checked against the 27 October 2025 version of the paper (SSRN listing confirmed through its DOI record; the full text was read from a copy hosted elsewhere). - Chandar & Klein Teeselink, "How Does AI Change Labor Demand? Evidence from 41 Countries," Stanford Digital Economy Lab working paper (September 2026)
Job postings and LinkedIn-based employment records across 41 countries, adoption inferred from generative-AI job ads, instrumented event study: affiliates of adopters have a junior share 1.9 points lower, driven by 6.7% senior growth, with junior employment not significantly lower. Uses affiliate-level headcounts by seniority, not individual worker histories; no analysis of promotions or training in the text read. - Danmarks Nationalbank, "AI Adoption and Firm Size: Employment Effects in Monthly Administrative Data," Working Paper No. 223 (16 September 2026) Judged from an abstract, summary or search listing onlySome details not confirmed
Landing page only read (authors not listed there): Danish firm AI adoption linked to monthly employer–employee records; 2023 adopters grew employment 11% less by late 2025, via lower recruitment, concentrated in smaller firms. The page does not mention worker age, trajectories or wages; it shows the data linkage this design needs exists. - Friebel, Huang, Li, Shukla & Zhang, "Pyramids, Diamonds, and Oscillations: AI and the Structure of Internal Labor Markets," NBER conference paper (10 April 2026)
Model distinguishing AI that raises junior or senior productivity from AI that changes junior learning; only the learning channel changes the long-run junior-to-senior ratio, and the channels have different wage effects. Descriptive postings evidence only; no worker-level test.
Search queries used
AI adopting firms junior employees promotion wage growth internal careers linked employer-employee data generative AImultinational affiliates AI adoption seniority junior share 41 countries working paperDanish firms AI adoption 2023 employment growth recruitment lower 11% administrative dataAI junior workers on-the-job learning apprenticeship erosion skill formation empirical firms fewer juniors trainingqualitative study interviews generative AI junior lawyers OR junior developers OR junior consultants mentoring training reorganisation of entry-level workAI adoption firms junior workers promotion wage growth career progression employer-employee data 2026 working paper
Provenance
RO-LD02-04 v1. Drafted by Claude (SocioLens Project Lead), AI-drafted from the LD-02 debate on 2026-10-12. Checked by an independent AI reviewer with fresh context (not a person), 12 October 2026: every nearest-work citation checked against its source page; corrections applied. Not yet reviewed by a domain expert. This is a research hypothesis, not a finding.