Originally published as a LinkedIn article by Steve Hyde
The Vanishing Entry Point
At Push, we're at the coalface of AI intgration across our client base. We see daily how quickly it's automating tasks, saving enormous amounts of time, and delivering exceptional results. The tools are already here, and they're exceptionally good.
The traditional agency pyramid, a wide base of entry-level executives doing admin-heavy work, topped by a thin layer of senior decision-makers is collapsing. We once had over a dozen account executives. Now it's just a few. Not because we're cutting corners, but because AI handles the repeatable work humans used to spend hours on.
The shape changed because the work changed. Clients don't pay for hours spent pulling data anymore. They pay for the judgment that turns insight into growth.
The Data is Alarming
This isn't just anecdotal. The collapse in graduate-level opportunities is measurable and accelerating:
The scale of the crisis:
- In the 2023/24 academic year, 1.2 million UK graduates competed for just 17,000 entry-level jobs, the worst ratio since records began in 1991
- Graduate-level job advertisements in banking and finance were down 75% in June 2025 compared to the same month in 2019
- UK tech companies cut graduate roles by 46% from 2023 to 2024, with a projected further 53% drop by 2026
- Youth unemployment (ages 16–24) rose from 10.9% in April 2022 to 14.3% in 2025
The Big Four consultancies have already scaled back:
- KPMG cut graduate hiring by 29%
- EY reduced intake by 11%
- PwC cut by 6%
And here's the most damning statistic: only 13% of graduate schemes include any AI training. We're sending graduates into an AI-transformed workplace with virtually no preparation.
From Pyramid to Diamond: How AI is Reshaping Businesses
Push has redesigned our entire structure into what we call a "diamond" shape:
- Fewer entry-level roles performing repetitive tasks that AI now handles
- A stronger mid-to-senior layer of strategists who think through AI, not just use it
- Leadership focused on governance, training, and commercial outcomes
This isn't unique to us. Across marketing and tech, rising employment costs combined with AI capabilities are creating a double whammy effect.
Historical precedent warns us: the 1980s National Insurance surcharge led to a 5% drop in manufacturing jobs; the 2011 VAT increase cost the UK economy an estimated 320,000 jobs. AI is already transforming how we work. When you layer cost pressures on top of automation capabilities, the incentive to reduce headcount intensifies.
The Institute of Student Employers (ISE) forecasts a 7% reduction in overall graduate hiring for 2025/26, driven primarily by sharp cuts from large employers. Meanwhile, the World Economic Forum reports that 40% of employers expect to reduce their workforce where AI can automate tasks.
This is structural change, not a temporary downturn.
The Graduates Who Will Thrive
Recently, we took on an intern who's still in his final year at a top university. He approached us directly, and here's what made him stand out: he wasn't studying computer science or anything remotely technical. But he demonstrated curiosity about AI and showed real examples of how he'd already used it.
Despite working only a few half-days per week, he's already adding significant value to our team.
That's the difference. Not the degree subject. Not the institution. The willingness to engage with AI now, learn it deeply, and apply it practically.
According to recruitment data, one in four recruiters now ranks AI as the most valuable skill for pay or promotion. Yet traditional university programmes in the UK have yet to integrate AI capabilities into their general curriculum.
The graduates succeeding in 2026 won't be those with the most polished CVs. They'll be the ones who understood early that AI fluency is foundational, and acted on it while still studying.
A Personal Wake-Up Call
My own son, Zac, turns 18 in a few weeks. He's studying Maths, Economics, and Theology at A-level, with offers from top universities to study management science.
I've encouraged him to seek out courses that help him understand AI's impact on business. But honestly? If the opportunity existed, I'd encourage him to go straight into a strong start-up to gain experience now, while learning and implementing AI processes.
There's a nagging fear I can't shake: the boom in this sector could be missed entirely by those spending the next three years at university.
I don't say this lightly. Education matters. But the pace of change is breathtaking. Stanford's Digital Economy Lab found that in occupations heavily exposed to generative AI, early-career workers (ages 22–25) have already seen a 13% relative decline in employment.
By the time today's freshers graduate in 2029, the workplace will look unrecognisable.
If you have children still young enough to choose their path, I'd implore you to consider the impact AI will likely have on their chosen field. Whether it's Law, Finance, or Medicine, be the one in your class who knows most about AI. It's no longer optional.
What Companies Can Do
At Push, we're actively looking at how we can take on graduates to give them a chance to gain proper work experience in a fast-paced, AI-driven environment. Can we justify it commercially? Not really. But we think more companies like ours should think hard about how they help out and give these young people a start in life.
We're creating the future workplace. We can't lock the door behind us.
Here's what businesses can do right now:
- Create AI-integrated internships and graduate schemes - don't just hire for traditional roles; build programmes that teach AI application from day one
- Partner with universities - 88% of organisations rely on informal on-the-job training instead of structured education. That's a gap companies and universities should close together
- Mentor and invest - even if you can't justify a full-time hire, create opportunities for project-based work, mentorship, or consulting arrangements that give graduates real-world experience
- Rethink job requirements - only 5% of employers still require a traditional degree for new hires, favouring technical AI certifications and bootcamp credentials instead. If AI skills matter more than a three-year degree, adjust your hiring criteria accordingly
The traditional entry-level job is disappearing. But the need for smart, adaptable, AI-literate talent has never been higher. If we don't create pathways for the next generation, we risk leaving them stranded, and simultaneously creating a talent shortage we'll pay for later.
Ready to see where your organisation stands on AI integration? Book a free AI marketing audit to understand how AI can strengthen your team, not just replace headcount.
About Push
Push Group is a UK AI-first digital marketing agency that has been embedding AI into marketing operations for three years. We help ambitious businesses go from AI interest to AI implementation, with measurable results from day one.
Learn more about our approach to AI in marketing or get in touch to discuss how we can support your team.
The Discussion: Industry Perspectives
Since publishing this article on LinkedIn, it's sparked significant discussion across the marketing, tech, and education sectors. Here are some of the most insightful perspectives from industry leaders:
On the Structural vs Skills Debate
Lee Tumbridge, CEO of Active Mentor, raised an important structural question:
"Most of this thread is treating a structural labour-market problem as a skills problem. 'Learn AI' is not the answer if AI has already removed the economic case for junior roles. The real question is simpler: who now pays for the years in which judgement, customer understanding, and commercial discipline used to be formed?"
point is critical. While AI fluency is essential for individual employability, it doesn't solve the systemic problem: if firms remove the entry layer entirely, they're not building innovation, they're consuming the pipeline that used to produce future operators at scale.
On Alternative Pathways
Fiona Clark, a career coach specialising in graduate employment, offered a practical alternative:
"I genuinely think short, targeted courses in AI could help many people get hired faster than taking a traditional degree route. A school leaver who uses those three years after school to build in-demand AI skills, gain hands-on experience, and develop a strong portfolio could, in some cases, be more job-ready than a graduate."
Paul Kearney, who focuses on AI implementation at scale, went further:
"AI isn't simply disruptive, it's constantly evolving disruption and I simply don't see how current Universities can service this evolving requirement. A technical apprenticeship in AI Implementation seems like a very solid job choice to me."
On the Human Skills Gap
Jo Scalpello, with experience across Nike, EA, and Trafalgar Entertainment, highlighted the human cost:
"If we don't have people learning the business, being exposed to problems and dealing with humans in (potentially tricky) buying situations then how do we serve a human-paying audience in the future? They are not training the leaders of tomorrow for their business."
Jo's observation about Apple's retail strategy, over-investing in customer service teams, is particularly relevant. In an age of automation, human-to-human contact becomes a differentiator, not a commodity.
The Entrepreneurship Angle
Philippe Masson suggested that the most critical skill for young people won't just be AI literacy, but time and energy management, entrepreneurship:
"For young people like Zac and my daughters, the most critical skill won't just be AI literacy, but high-level time and energy management. AKA Entrepreneurship!"
When I look back at my career, it was always the mavericks who tried something risky that I looked up to. Perhaps that's the mindset graduates need now.
A Longer-Term Comparison
Howard Veary, retired Deloitte partner and former Governor at Wellington College, offered perspective from the 2008 financial crash:
"Much worse for new graduates entering the job market was the immediate aftermath of the 2008 financial crash. My daughter took herself to Australia and stayed for 4 years and got herself on the career ladder. If you are determined and resourceful, you will succeed."
I appreciate Howard's optimism, but my prediction is that this erosion of entry-level jobs will be deeper and last longer than the 2008 crash. We're still in the early days of AI's impact on graduate employment.
What are your thoughts?
Are you seeing similar shifts in your sector?
Share your perspective on LinkedIn.
