Get Quote
13%

Employment among 22–25-year-olds in AI-exposed jobs is down 13% since 2022. (Stanford Study)

32%

UK entry-level vacancies down 32% (Adzuna)

underline

Fresh graduates are fighting for visibility in a flooded market with few roles. The UK’s Big Four have slashed junior intake by as much as 29% as they outsource tasks to AI. Junior software developers, once a hot commodity, have been struggling to find their first roles in a market that demands two years of experience.

The rise of AI isn’t necessarily the root cause — but it’s the thing that tipped the scales. And if organizations continue to pass over the most junior members of their workforce, they’ll only set themselves up for an uneven workforce in years to come.

Here’s what’s happening — and how talent leaders are adapting their hiring strategy for junior hires in an AI-shaped market.

Why is early career hiring on the decline?

Since the release of ChatGPT, early career hiring has taken a backslide. A 2025 Stanford study found that workers aged 22 to 25 in roles most exposed to AI have seen a 13% decline in overall employment since 2022. Some of those roles include customer service representatives, software developers, and accountants. Across the pond in the UK, a 2025 report from job site Adzuna found that vacancies for entry-level roles have fallen by 32%. 

It’s a reality that’s already having a consequence on early career workers.

“Right now, we’re on this path where the present entry-level workers are being boxed out of the market,” said Tracy St. Dic, Head of Talent at Zapier on the first episode of Talentful’s The Debrief podcast. “Because AI can accelerate the learning curve of knowledge, early career jobs as we know it are not going away — we just need to redefine our assumptions on what the talent can do. The path I’m more optimistic to see is that we’re no longer going to assign the grunt work to entry-level workers, which hopefully will then empower them to start creating value [for] the company from the very beginning.”

To St. Dic’s point, the problem isn’t that organizations need fewer early career hires — it’s that companies are operating on outdated assumptions about what junior talent needs to learn and contribute in their first few years in a role.

In the pre-AI world, the talent acquisition model was simple: hire someone cheap, ramp them slowly with entry-level tasks over the course of 18 months, and eventually they’d absorb enough to move up. But now, with AI able to take on many of these tasks, the question becomes: what’s left for junior talent to do?

“We assumed that early level roles had to do grunt work to learn and that’s just not really the case,” St.Dic said. “It was advantageous for more experienced workers to assign that type of work to junior staff… But it’s not inherent that those things have to be done in order to learn.”

So what happens if companies don’t get the memo quickly enough?

“All of the companies that are devaluing graduate hires and what they can achieve in a company, it’s just going to be maths,” said Chris Abbass, CEO of Talentful. “At some point they’re going to have a top heavy organization. They’re not going to have enough people to do the work they need them to do. They’re going to run out of talent.”

How to hire for early career talent now: 4 strategies

The future of talent isn’t about sidestepping early career hires — it’s about rebuilding the model for how junior hires learn and contribute in an augmented world.

That means rethinking not just what you’re hiring for, but how you’re evaluating for it in the first place — and how you’re setting up early career talent for context-building, curiosity, and succession.

1

1. Drop the experience requirements that don’t matter.

The traditional recruiting model has always hinged on the premise that you need to have experience to get experience — it’s a neverending cycle. And as AI takes on those early career experience-building foundational tasks, it’s not leaving much space for junior hires to earn their stripes.

“We assumed that early level roles had to do grunt work to learn and that’s just not really the case,” St. Dic says. “You don’t need to be a coordinator on a recruiting team for two years before you can become a sourcer. You don’t have to be a sourcer before you can become a recruiter.”

If early career experience is no longer the proving ground it once was, then we need a new lens for spotting potential. Instead of screening for time served, we should be assessing how someone thinks, learns, and grows. The question isn’t “Have they done it before?” — it’s “Do they have the capability to do it now, and the capacity to get better fast?”

What to look for:

  • How fast can they pick up new concepts?
  • Can they break down complex challenges into manageable parts?
  • Can they explain their thinking and decisions?
  • Do they ask questions, iterate, and work with input?
  • Do they understand core concepts (not just specific tools)?

2

Screen for AI fluency, not AI skills.

By this point, almost anyone can generate an output with AI. But the difference between hiring for AI skills and fluency lies in how early career candidates use the tools to get to a conclusion.

“There’s a huge difference between somebody who takes voice of the customer, sales deck, some market research, throws in AI and says, here’s a positioning document for this product… and someone who looks at it and they’re like, this is 20% there. And I need to edit this whole thing or add more context and can tell the difference between good and great,” said Ben Sesser, CEO and co-founder at BrightHire.

This judgment is where talent will naturally rise to the top. Some organizations are already building it into their talent acquisition process.

Case in point: Consulting magnate McKinsey added a new stage to its hiring process for final-stage candidates that assesses their ability to work with and prompt an AI. The task sees candidates prompting an internal AI to complete exercises resembling real client scenarios — with an emphasis on how candidates use the system and their reasoning skills to achieve their goal.

The goal isn’t to find candidates who know how to use AI effectively — it’s to look beyond just skills and focus on how people work and learn.

 

“What’s their capacity to learn, capacity to take on new knowledge, capacity to work with others and collaborate effectively?” said Matt Bradburn, founder of PeoplexAI, in our recent article on AI skills gaps. “What’s their kind of mental elasticity? Can they grow and develop that over time, or are they going to simply hit a level and level out?”

Give candidates a task and make it clear they can use AI. Then ask how they prompted it, what they didn’t trust, how they verified it, and what they changed. The strongest people treat AI like a junior assistant — helpful, fast, full of potential — but never beyond supervision.

What to look for:

  • Can they explain their prompting logic and what they asked the AI to do?
  • Can they identify what needs verification or might be wrong in the AI’s output?
  • Can they describe how they validated the response (sources, tests, cross-checks)?
  • Can they articulate what they changed and why?
  • Do they treat AI as a tool that needs supervision, not an authority to trust blindly?

3

Look for builders, not credentials.

For decades, the resume has been king — a highlight reel of where a candidate has been and what they’ve done. But it’s really more of a rearview mirror in today’s market — which is why talent teams are turning their attention to what candidates are actually doing right now, not their polished trophy cabinet from years gone by.

Design portfolios, Github projects, side hustles and personal passion projects unearth the real doers — the people who build because they’re curious, not just because it was part of their quarterly KPIs.

“People really need to show up and show their work,” said Sesser. “Folks can come off a college campus and be like, look at all these things I built. This is my evidence. This shows that I’m AI native, that I can build, that I get stuff done.

“There’s going to be a huge delta between early career folks that are high agency, high curiosity, and super hungry and the economic journey they’re going to go on throughout their career and others,” he added. “Because if you do have those three things, you will take full advantage of an incredible set of tools and just completely differentiate yourself.”

 

Evaluating that evidence in a consistent and structured is where talent teams need to hone their muscle when hiring early career talent.

What to look for:

  • Can they explain not just what they built, but how they approached the problem?
  • Can they articulate why they made specific choices?
  • Do they walk you through where things broke and what they learned?
  • Can they articulate the trade-offs in their decisions?
  • Do they show self-awareness about their capabilities and limitations?
  • Do they focus on their thinking process, not just the final output?

 

 

4

Design intentional learning systems that start from onboarding.

Dropping the years-of-experience requirement on a resume solves part of the early career hiring problem. But it creates a new one: How do early career hires get up to speed and build context when AI can do all the grunt work?

“AI doesn’t automate away jobs, it automates tasks,” St. Dic said. “The tasks that people are doing are going to have to change. Companies will have to design more intentional learning systems and learning ecosystems where AI is not a replacement for practice, but practice really looks different,” she added. “I see a world where there needs to be more mentorship, more shadowing, [and] more scaffolded learning so that people can build judgment and problem solving and understand the context that they used to get by just watching the work.”

 

 

Some organizations are already putting this into practice with junior hires.

“We have shortened the ‘phase of mechanical learning’ and are moving to context faster: product, users, solutions,” said Tetiana Hnatiuk, Head of HR at Skylum Software. “AI [completes] simple tasks, and we teach you to think. Now onboarding has become more mentoring. Junior professionals are previously involved in real projects, but with clear frameworks and support — this accelerates growth.”

Building the early career hiring model of the future

As the race to conquer AI continues, junior hires are organizations’ biggest bet for future succession planning that sidesteps the issue of an imbalanced workforce. 

But it won’t be enough to wait for the market to correct itself. Future-thinking organizations will be seizing the opportunity to redesign the early career hiring model — creating new ways to hire, engage, and define an ecosystem that rewards curiosity, fluency, and judgment over years of experience. 

In practical terms, this will involve reframing the entire experience for junior employees — from the tasks used to define their roles and how they’re assessed for raw skills like learning and curiosity, to the structure of how they actually build context through onboarding. 

Ultimately, the organizations that move forward together with AI and people won’t be the ones with the smartest machines — they’ll be the ones focused on molding the smartest people. That has to start with the people they hire, and how they help them grow.

Early career hiring needs to evolve — and so does the way you recruit it. Talentful’s embedded RPO teams help you hire smarter, faster, and with long-term impact. Get in touch to see how we can support your early talent goals.


Speak to our team