Early Careers “crisis” - it’s not all about AI

In the 1960s, two researchers walked into a primary school and told the teachers they had a test that could spot which children were about to bloom. They pointed out the handful who were, supposedly, on the edge of a sudden leap. It was not true. The names had been pulled out of a hat. But when the researchers came back at the end of the year, those children really had surged ahead, and not because anything about them had changed. It was because of how they had been seen, and spoken to, and expected to do well, once the adults believed they were destined for it. Psychologists borrowed a name for it from the myth of the sculptor who loved a statue into life, and called it ‘The Pygmalion effect’ - we tend to become the story we are told about ourselves.

I want to start there, because right now, as a country, we're at risk of running that experiment in reverse on a whole generation. Open a paper, scroll a feed, and the message is relentless and roughly the same. The jobs are gone. AI ate them. A generation is on the scrapheap. So it's worth asking, before we repeat it one more time, whether the story is even true, and what the telling of it is doing to the people who hear it.

Because there really is a door here, the entry point into work, and it really is changing shape. You did your GCSEs, maybe a degree or an apprenticeship, you knocked, and for most people it opened onto a first job with a bit of room to find your feet. Look closely now and it's narrowed on several sides at once, and only one of those sides is really about AI.

The ground really has moved

In April to June 2026 there were 981,000 sixteen to twenty-four-year-olds in the UK not in education, employment or training, which is 13% of them.

Of that 981,000, some 393K are unemployed and 588K are economically inactive, so six in ten aren't looking for work at all. That isn't a technicality. Unemployed means you're looking and haven't found it. Economically inactive means you've stopped looking. Ill health is the single most common reason young people give for being NEET (DWP). The Work Foundation reads the same shift and says it “suggests a new level of disengagement with job seeking”, which is a careful way of saying this is no longer only a shortage of jobs causing the problem.

The story we're telling is only half right

The dominant line is that AI has eaten the entry-level jobs. Look at the hiring data rather than the headlines and that's not quite what's happening.

Entry-level hiring has fallen broadly in line with hiring overall, and there's no separate collapse in first jobs that AI has clearly caused (LinkedIn's UK Labour Market analysis). The gap only opens inside the occupations where AI is reshaping the work itself, and even there LinkedIn is careful to say it doesn't yet know why.

The longer trend should worry us more. Starter jobs, the ones you can walk into without experience, have fallen by 49% over the past decade, from a weekly average of 139,329 vacancies to 71,110 (Work Foundation).

And the collapse isn't evenly spread. Sales and customer service fell 59%, and inside that, sales and retail assistant jobs dropped by 53%. Those are the classic first rungs, the shop floor and the call centre, the jobs that taught many of us how work works. They didn't go because of a language model. They went with the high street, the cost-of-living squeeze and a decade of thin growth.

Two other forces are at work too, and neither is AI either. Fresh LSE research separated out AI and remote working and found it's remote working that robustly predicts the fall in early-career hiring, with the AI effect all but vanishing once both are in the model. The reason they point to is that remote work raises the cost of supervising a junior and slows the learning that comes from being near people. And more than half of Gen Z say they can't afford to own a home, with around seven in ten saying the availability or cost of housing is shaping their career decisions (Deloitte). So you get a generation that can't easily move to where the entry jobs are, arriving at a doorway that had already narrowed long before any of us typed a prompt.

Where AI does bite is worth being precise about, so we don't overclaim it. It isn't deleting the first rung, it's raising the height of it. PwC went through US job adverts and found the most AI-exposed entry-level roles are now seven times more likely than the least exposed to ask for skills that used to belong to senior people. Motivational leadership. Team building. Stakeholder management. Data-driven decision making. PwC sets that list under a question I'd pass straight on to you: did your first job require these skills? In those most-exposed roles, senior capabilities now account for 52% of the new skills being asked for, against 7% at the least-exposed end. The routine tasks that used to be how somebody learned the ropes are the ones being automated, so what's left of the job asks for judgement, sooner.

Employers are telling each other the same story, too. When Survation polled 1,001 UK senior business leaders for the Work Foundation in August, 36% said they had cut entry-level jobs for 16 to 24 year olds in the past twelve months, and it's the bigger employers doing it: 48% of medium-sized and 46% of large, against 24% of small. More than two in five, 43%, said they had invested in AI or automation that reduced those roles, rising to 60% among large employers. Put that next to the LSE finding. Employers believe AI did it. The stronger evidence points at remote working. Both of those are true at the same time, and the belief is doing work of its own, because what you think took the jobs away shapes what you do next.

The last thing pressing on the door

There's one more thing pressing on that doorway and it isn't structural at all. It's the story we keep telling about it. We've just seen the data doesn't support the strongest version of it. But the story does its work anyway, and to see why, it helps to know what hope is actually made of.

Not the greetings-card kind. The psychologist C.R. Snyder spent a career showing that hope is really three things holding hands: a goal you want, a believable route to reach it, and the sense that you're able to walk it. Take away the route, or the belief you can walk it, and hope doesn't dim, it collapses. Which is what a narrowing doorway does from two sides at once, and what the doom narrative piles on top.

A message repeated at that volume doesn't stay outside the head. It's the oldest finding in media research, that the stories we're told shape not just what we think about but how we're taught to feel about it. It's the classroom again, the Pygmalion effect we opened with, only pointed at a whole generation instead of a handful of children. A young person told often enough that the door is shut stops knocking, and the not-knocking is part of what keeps it shut.

And we can see it landing. When Handshake asked 10,678 UK students and graduates how they felt about the graduate job market this year, 65.7% called themselves pessimistic and just 15.4% said they were genuinely optimistic. Nearly half expect to apply to twenty-one roles or more, and more than a quarter expect to send over forty. It's worth being clear who that sample is. These are the ones still trying, still in the pipeline, still filling in the forms. The 588,000 who have stopped aren't in it at all. If even the ones still knocking are two-thirds pessimistic, the story has already done its work.

So, what's ours to do?

Three things then. Most of us can't fix the housing market, or the economy, or the headlines. But hope has its antidote built into its own definition, and it's one employers are unusually well placed to supply. If hope is a believable route plus the belief you can walk it, then the two most useful things we can offer are a visible route and living proof that people like them have walked it. There's a third as well, and it's closer to home than either.

And we don't have to guess at what would help. Asked what mattered most to them in a first role, those same students and graduates put learning new skills first at 31.7%, career progression second at 26.6% and mentorship and support third at 16.5%. That's nearly three-quarters of stated preference, and all of it is development. Competitive salary came fifth, at 8.2%. Which isn't to say pay doesn't matter, it plainly does, and nearly seven in ten of them say a missing salary in an advert makes them less likely to apply, or stops them applying altogether. Pay is the gate. Development is the reason they walk through it.

The first, the route, is the easy part to name and, too often, isn’t visible. M&S made a move worth borrowing from, a thousand training places across its stores over eighteen months with no degree required, six months in-store and then further training towards store manager, on the plain principle that you need attitude and a willingness to learn more than a certificate. That's a pathway made visible, which is half of hope handed straight over. Apprenticeships can do the same, though left to drift a large share of apprenticeships still go to older, existing staff, so if early careers is the point it has to be designed in, not assumed. My time at BAE systems made this abundantly clear and they do apprenticeships employer branding exceptionally well. Yes, they’re in all the channels you’d expect, Higherin, which used to be Rate My Apprenticeship, GetMyFirstJob, the government's own Find an Apprenticeship service, but they go way further – partnerships with local schools and colleges close to their sites, online apprenticeship discovery sessions, and what is often missed, activity designed to support parents in guiding their kids around there post education options.

The proof is the other half, and it's the half employer brand and advocacy were built for (not a new job for us, just the job we already do, pointed somewhere it matters more than usual). Not a glossy graduate campaign, but the honest story of someone who came in on the bottom rung and got somewhere, in their own words. Again BAE do this brilliantly – a huge range of colleague stories from current and past apprentices that they surface and spotlight. Showing truly diverse backgrounds, so that an apprentice can see someone like them that went before them. That is instilling that belief you can walk into it at granular level.

And then there's the third, which is the only part of this entirely ours. When the Work Foundation sat down with young people who weren't in work, education or training, what came back wasn't mainly about the economy. It was about applying. Platforms that won't let you build a CV if you haven't got qualifications, so you're out before you've started. Adverts still up for roles that went weeks ago. Not knowing whether a person, a filter or an algorithm read you. And the one every single participant raised: no feedback, ever. What that does is the part that should stop us. Young people who'd had a generic rejection and no explanation had, in the researchers' words, "internalised the outcomes as a personal failure", not recognising it as a feature of the system rather than a verdict on them. One 20-year-old put it as "I keep thinking what have I done wrong".

That isn't a labour market problem. That's our processes. It's worth walking your own application journey this week and getting to the end of it as if you were eighteen, with no qualifications and no data allowance, because that journey is telling them something about themselves, and your employer brand is saying it whether you wrote it or not.

The doorway is changing shape, and most of why is out of our hands. But the story a young person hears about whether it's worth walking up to at all, and whether there's a way through once they do, that part we absolutely can and should write. Every job ad, every careers page, every colleague story is a small vote for one version or the other. It's the Pygmalion classroom again, and this time we're the adults in it.

And if you fancy a chat to explore how I might support your organisation with your EVP or perhaps a TVP (smaller version for one talent group like early careers), just drop me a line.

Vicki Saunders

Founding Director, The EVP Consultancy

With over 19 years in employer brand and EVP strategy, Vicki works with organisations to build employee value propositions that are honest, distinctive, and have impact. She's the creator of the 8-Dimensional EVP Framework and The EVP Edit newsletter.

vicki@theevpconsultancy.co.uk