Leadership1 publisher2 min readPublished
Firms that adopt AI are adding senior staff while their junior ranks shrink
Stanford and King's College London researchers found senior roles at AI-adopting firms grew 6.7% over five years while junior roles fell 3%. Junior development is now a leadership decision, made on evidence that so far amounts to a newsletter summary of one paper.
The Board Room · Leadership desk
What happened
- Overall employment at the firms that adopted AI rose 3.3%, according to the same research.
- A separate study found that workers with greater access to information are less likely to remember it, though that paper did not directly study AI.
- Unions have made AI protections a top bargaining priority, seeking notice of new technology, human oversight of AI reports, and training and re-skilling.
Compiled by The Board RoomSomething wrong?How this is made
Why it matters
- decision With seniors rising about 10% relative to juniors, firms must choose how to fill future senior seats: promote from a smaller internal pool or buy experience from outside.
- cost Fewer junior roles lower payroll now, and the cost comes later as fewer internally trained candidates when senior roles open.
- exposure Juniors who rely on AI to retrieve knowledge may retain less from their work, weakening the on-the-job training the remaining junior roles are meant to provide.
- constraint Union demands for re-skilling clauses could turn junior training into a contractual obligation wherever firms bargain collectively.
On those figures, the number of senior employees for each junior one rises by about 10% over five years [1]. The calculation is 1.067 divided by 0.97, and it holds whatever the starting mix was. A firm with four seniors for every ten juniors ends the period with roughly 4.4.
The board-deck version is that AI adoption comes with net hiring. Employment at adopting firms rose 3.3%, according to the Stanford and King's College London researchers [1]. That line is accurate but leaves out succession. The growth sits in senior roles [2], and the junior roles that feed them lost 3% [3]. A junior the firm does not hire this year is someone it cannot promote from within later in the decade.
A skeptic would say this is one paper, seen through a newsletter summary, and that a 3% decline over five years is modest [3]. Both points are fair. The summary does not name the authors, describe the sample, explain how roles were sorted into junior and senior, or show that AI caused the split. I'd still treat it as a planning input, and I'd revise that view once the methods are public. Promotion pipelines take years to refill.
The remaining juniors may be the harder problem. A separate study of what it calls the connectivity tradeoff found that workers with greater access to information are less likely to remember it [5]. That paper does not study AI directly [6]. Its researchers still warn that employers who encourage staff to rely on AI to retrieve knowledge may be "optimizing for access while unintentionally undermining learning" [7]. The counterweights they suggest are applying information to novel problems, discussing it with colleagues and teaching others [8].
Where workers bargain collectively, management may not get to make this choice alone. Unions have put AI protections at the top of their bargaining agendas. Negotiators want training and re-skilling commitments, along with notice of new technology and human oversight of AI-generated reports [4]. A re-skilling clause in a contract would make junior development an obligation instead of a budget line that can be cut.
This quarter, the decision is how many junior seats to keep and what those people do all day once software handles retrieval. The consequence comes when a senior role opens. At that point the firm chooses between an internal candidate who trained with fewer peers and less hands-on work, or an outside hire paid at senior rates.
What to watch
- Publication of the full paper's methods: its sample, how it defines junior and senior roles, and whether it tests if AI caused the split.
- Whether signed union contracts begin to include AI training and re-skilling clauses.
- Later data showing whether junior roles at AI-adopting firms keep falling beyond the five-year window.