Views and announcements

In the age of AI, cutting junior engineers risks a decade-long skills gap – why Kainos is investing instead

  • Aislinn McBride, CTO at Kainos, discusses the growing risk of a decade-long skills gap due to significant cuts in hiring junior AI engineers despite ongoing demand for senior roles.

    There's a shift happening in the tech labour market that feels rational in the short term but risky in the long term.  

    At Kainos, we are choosing a different path. We continue to invest in early careers, deliberately shaping the journey from school leaver to engineers contributing to production, building skills that last a lifetime. We have seen first-hand how sustained investment in junior talent fuels long-term growth, while also creating meaningful social impact across the regions we operate in. 

    Hiring has slowed. That's expected. What's different this time is where the cuts are landing. Entry-level roles have been hit hardest, while demand stays concentrated at senior levels. [nfer.ac.uk] 

    READ MORE: Yesterday’s Defences, Today’s Attacker

    We've seen versions of this before, when hiring contracts sharply, early-career intake is usually the first thing cut and the last thing restored and the cost doesn't show up on the balance sheet that year. It shows up later, in the experienced engineers who never came through. 

    On the surface, the logic holds. When pressure rises, organisations optimise for immediate productivity. They hire experience, reduce training overhead, and increasingly lean on AI to fill the gaps. But this isn't only cost control, it's a decision about how our industry sustains itself. 

    The signals are already pointing one way 

    The data across the UK and US tech market is increasingly consistent: 

    • In US Big Tech, new-graduate hiring has fallen by more than 50% since 2019, with graduates now making up just 7% of hires. [SignalFire] 
    • In the UK, entry-level vacancies — graduate roles, apprenticeships and junior positions — are down roughly 32% since late 2022, with IT among the hardest-hit professional sectors (down around 55%). [Adzuna / The Independent] 
    • In the US, early-career workers aged 22–25 in the most AI-exposed occupations have seen roughly a 13% relative decline in employment since 2022 – though this research is recent and still debated. [Stanford Digital Economy Lab] 
    • No single number tells the whole story, and the drivers are mixed — AI alongside higher employment costs, the end of cheap capital, and a correction after the 2020–22 hiring boom. But taken together, the trend is hard to miss: the entry point into the profession is narrowing. [techuk.org] 

    This would be manageable if the system were healthy 

    It isn't. The system is already under strain. 

    • Employers continue to report difficulty recruiting the skills they need. [theiet.org] 
    • A significant share of experienced engineers are approaching retirement. [manpower.co.uk] 
    • Apprenticeship pipelines are weakening, with fewer starts and completions in key disciplines. [engineerlive.com] 

    This is not a market with surplus talent. It's a system already running close to its limits — and reducing the intake into it doesn't stabilise it. It increases fragility. 

    AI is changing the pathway, not removing the need 

    AI is often cast as the counterbalance, and in some ways it is. It reduces the need for routine, entry-level tasks; boilerplate code, data preparation, simple analysis. But those tasks were also how engineers learned. They built intuition through repetition, developed judgement through failure, and came to understand systems by working through their edge cases. 

    When those tasks disappear, the learning pathway doesn't automatically evolve to replace them. Left alone, it simply breaks. 

    It's worth taking the strongest counter-argument seriously: perhaps AI means we genuinely need far fewer junior engineers, permanently. I don't think the evidence supports that. The work juniors grow into judgement and system design; based on a solid understanding of technology  is rising in value, not falling. The risk isn't that we need fewer engineers. It's that we invest less in how engineers become experienced, while still expecting the same outcomes at scale. 

    The effects show up inside organisations sooner than we think 

    The impact of reducing junior hiring isn't immediate. In the short term, almost nothing appears to change. Delivery continues, experienced hires cover the critical roles, and AI lifts throughput. 

    But the effects accumulate: 

    • In 3 years:  fewer engineers ready for mid-level roles 
    • In 5 years:  thinner leadership pipelines 
    • In 7–10 years:  structural shortages of experienced talent 

     And before those timelines fully materialise, the pressure starts to show up inside organisations: 

    • The lack of people growth because they need to keep doing the same work because there’s no junior coming behind them. 
    • High attrition due to lack of opportunities to lead others or develop themselves.  
    • Decline of leadership skills because seniors have had no-one to lead and mentor. 
    • Culture of the organisation as the short term gain is prioritised over developing people. 
    • Rising cost model of only employing experienced staff. 

    By the time it's visible, it's  a systemic constraint and when every organisation optimises the same way at once, the pattern is predictable: a workforce that skews too senior, a shrinking pool of promotable talent, rising competition for experienced engineers, and rising long-term cost to hire and retain them. 

    Skills gaps form when many reasonable decisions are made in parallel, rather than from one bad one. 

    How we design the pathway should be the question we ask 

    Entry-level roles should evolve. But removing them altogether isn't a strategyrather the absence of one. The organisations that navigate this well will be more deliberate: 

    1. Maintain intentional entry points.  At a high enough level to sustain long-term capability,  not a token intake.
    2. Redesign early careers around capabilityLess repetitive task-work, more systems thinking, problem framing, and working effectively with AI from day one.
    3. Make development explicit, not incidentalIf experience is no longer built through repetition alone, it must be designed; through mentoring, exposure, and structured progression.

    At Kainos, we've always invested in early careers. Our entry-level academy pairs an intensive four-month training programme with a year of structured learning alongside real project delivery. By aligning that model to how engineering is changing in an AI-enabled world, our early-career engineers contribute from day one on delivery and, just as importantly, act as a catalyst for capability-building across the organisation, spreading new tools and ways of working through the projects they touch.  In the year ahead, we’ll add circa 200 early careers to our workforce in Kainos. This sustained approach to early careers has led to a retention level in the 90s across the organsation, shaping not just our talent profile, but the culture that’s embedded across the organisation. 

    It takes more intent. It is far less costly than rebuilding a pipeline that's been allowed to collapse. 

    READ MORE: AICON 2026 Announces ICC Belfast as New Home for Ireland's Leading AI Conference

    We are making a long-term decision, whether we realise it or not 

    Talent pipelines run on a different timescale to market cycles. The choices we make now will shape the availability, capability and cost of engineering talent for the next decade. If we keep reducing junior intake without redesigning how engineers actually develop, we aren't simply responding to today's conditions. 

    We're setting the conditions for the next skills gap and those gaps are slow to emerge, but even slower to close. 

    Learn more about Data & AI at Kainos

    Read the Summer 2026 edition free online →

    Stay connected with NI's tech community:

Share this story