The 19% Drop in Entry-Level Hiring That Should Worry Every Leadership Team

You've probably noticed it already, it certainly isn't quiet on your socials. The graduate who used to sit in the corner of the office, learning the ropes by doing the tedious work nobody else wanted?

That desk is empty now.

A Stanford study published this month put a number on what many we've been seeing: employment among young workers in AI-exposed occupations has fallen 19% compared to roles where AI has less reach. A notedly measured drop, already recorded.

I've been turning this over since I read it, because the conversation about AI & jobs has been stuck in the wrong place. Most boardroom discussions focus on whether AI will replace senior professionals, whether it will write the strategy deck or make the CFO redundant. But the damage is showing up somewhere else entirely: at the bottom of the org chart, in the roles that were always treated as disposable but were never actually disposable at all.

Those entry-level jobs did something that no onboarding programme or e-learning module achieves. They taught people how an organisation actually works. How decisions get made in practice rather than in theory. How to read a room, when to push back, when to say nothing. The junior analyst who spent two years cleaning data before anyone let her build a model learned things about that data, about that business, that the model itself will never know.

When you automate those roles without replacing the learning they provided, you lose the pipeline. Five years from now, you go looking for a mid-career hire who understands your sector from the ground up, & that person does not exist. They never got the chance to become that person.

This is what L&D teams & talent leaders need to grapple with now. The Stanford numbers tell us the pathway from "new graduate" to "future leader" has a hole in it, & most organisations have not even noticed.

I talk to senior leaders about this regularly, & I hear two common responses. The first is denial: "We still hire graduates." Sure. But too often those graduates are managing prompts for tools they don't understand rather than doing work that teaches them how the business operates. The second is fatalism: "If AI can do the job, maybe we don't need the role." That logic sounds clean until you realise the role was building the person. The person was the point.

What I hear from senior leaders is that they haven't really thought about the long-term process & long-term workforce planning that needs to happen to keep their organisation running. They're thinking more about how they train AI rather than how they train their workforce with those skills. Most organisations need to cultivate that experience in their organisation to make AI an effective tool for them, because AI augments people's capability but it can't augment capability that hasn't been cultivated. You have to cultivate that capability first, & then you're upskilling the workforce as well.

A practical response starts with honesty about what entry-level roles actually achieved beyond their job descriptions. Then it means redesigning early-career experiences so that AI literacy is built into the work from day one. A graduate who learns to work alongside AI & develops real judgment about what the tool can & cannot do is a stronger hire than one who just did the manual version of the task. But only if someone designs that experience with intention.

One thing worth noting is that AI is an intent driven interface. Because of that, you need to help people with clarity of mind, with clarity of outcome in order to get the tools that you need. But it comes back down to communication as well. Communication is the most effective skill that you need during the digital era. If you can communicate what you want, you can create it. If you can imagine it, you can create it. At the basics level, it's almost going to be like having people join your organisation who don't understand how email works. If no organisation is cultivating these fundamental skills, then it's going to be a competition for talent over a skill that is quite easy to go & learn. All organisations share a responsibility for making sure that people have these hygiene skills. & if you don't provide that at an entry level, then those young people are going to struggle in mid-career progression.

The organisations that treat this as a hiring problem will keep cutting entry-level roles & wondering why their leadership bench looks thin a few years later. The ones that treat it as a design problem will build the kind of talent depth that no amount of late-stage recruitment can match.

My prediction: within two years, the companies that cut entry-level roles fastest will be the ones paying the steepest premium for mid-career hires, because they will have nobody coming through internally. The signal to watch is whether "leadership pipeline risk" starts appearing in annual reports & investor calls. I would change my mind if organisations prove they can replicate early-career learning through structured AI apprenticeship programmes at scale, but I have not seen credible evidence of that yet.

The Stanford data is a gift, if you use it. It gives L&D leaders the evidence to argue for investment before the gap becomes a crisis, & the best place to start is building real AI literacy from the ground up.

Frequently Asked Questions

How is AI affecting entry-level jobs right now?

Employment among young workers in AI-exposed occupations has already fallen 19% compared to roles less affected by AI, according to a Stanford study. This is a measured decline in existing hiring patterns. The drop is concentrated in roles where routine tasks can be automated, leaving fewer opportunities for graduates to enter the workforce through the paths that used to absorb them.

Why should leaders care about junior roles being automated?

Entry-level roles are where future leaders learn how organisations actually operate & develop judgment. Removing those roles without replacing the learning they provided creates a gap in the leadership pipeline. That gap becomes visible three to five years later when there are too few mid-career professionals with the depth of understanding your organisation needs.

What can L&D teams do to protect the talent pipeline?

Start by auditing what early-career roles actually taught people beyond their formal job descriptions. Then redesign graduate experiences so AI literacy is embedded from day one, with new starters learning to work alongside AI tools rather than being replaced by them. The goal is to build judgment & contextual business understanding together with technical skill.

Does AI literacy training alone protect entry-level workers?

AI literacy on its own is not enough. What protects early-career workers is understanding AI's capabilities & limitations alongside the contextual business knowledge that only comes from doing real work inside an organisation. Training belongs inside the job itself, embedded in daily responsibilities.

How long before the loss of entry-level roles affects leadership hiring?

The effects will become visible within two to three years as organisations that cut junior roles find they have fewer internal candidates ready for mid-career positions. The Stanford study's 19% employment drop has already been measured, so the pipeline consequences are already in motion. Organisations that act now have a window to redesign early-career pathways before the shortage hits.