Photo by RUT MIIT on Unsplash

Back in the day – by which I mean pre-2022 – company structures used to look a lot like a triangle. At the bottom sat the entry-level graduate candidates, making up the largest proportion of the employees, followed by a slimmer section of middle management in the stage above. Towards the top were the executives – an elite few – and the star at the very top of the triangle was the CEO.

Now it’s a diamond. AI has taken the triangle structure and filed down the bottom level, where the graduates used to perch, into a sharp point. Graduates are struggling – in 2026, the average entry-level graduate job receives 140 applications, according to the Institute of Student Employers, compared to around 86 applications per vacancy in 2022. This increased demand is driven by increased job seeking, AI replacing entry-level jobs, and ergo reduced vacancies.

Meanwhile, the middle of the diamond is jam-packed full of middle managers who started in entry-level roles and worked their way up. Where previously they might have moved on to bigger roles at other companies, they now find their progress stagnated; held back by fears of a difficult job market, economic strain, and worries that AI will replace their skills if they leave.  

That diamond shape might also be set to change again in the near future, squeezing the middle as AI continues to influence how people work. “Too few people at the bottom now means a massive skills shortage later,” warns Isabel Berwick, Financial Times journalist and Host of the Working It podcast. It’s the entry level roles which are usually where organisations develop future leaders, she says, and with these roles now being replaced by AI, these leaders are unable to develop the skills they need to progress.

Recruitment is broken, or at least severely damaged. But can AI be used to fix the problem it has created? 

Can AI ever be a fair recruiter? 

It’s an undeniable fact that most companies are already using AI in their recruitment processes. Recent data from DemandSage finds that 87 percent of companies are doing so, and two-thirds of recruiters say the main advantage of this is saving time.

Despite this, 35 percent of recruiters are concerned that AI may exclude candidates with unique skills and experiences, those elements that turn a great employee into an exceptional one, giving companies a better, more competitive edge. 

According to Standout CV, recruiters are using AI to write job descriptions (60%), read CVs (59%), source candidates (51%) and even perform interviews (20%). If LinkedIn is to be believed, people generally have negative opinions about AI interviewing them for job roles. 

At the start of this year, senior features writer at The Independent, Helen Coffey, trialled a TestGorilla AI interviewer, who she names Carl. Considering she is just trialling this software, Coffey reports feeling surprisingly nervous during the interview. 

“I realise my anxiety specifically stems from the fact that Carl is not a real person,” she writes. “I realise just how much I’ve always relied on my people skills to carry me through interviews. Even if I fudge an answer, I’m confident in the fact that those less tangible, “soft skills”  – emotional intelligence, the ability to make people smile or put them at ease with a well-placed joke – will go some way to making up the deficit.” 

She adds that normally – whether you’re interviewing online or in person -there is a positive feedback loop created by your enthusiasm, that the other person reflects. This isn’t the case with AI. 

Speaking of her AI interviewer, Coffey says, “I can tell his heart’s not really in it. After all, he doesn’t have a heart.” 

AI tools are being sold to companies as a productivity miracle. When AI is used as an interviewer, supporters argue that it can give a fairer overview of a candidate, rating the individual against the role’s specific skillset and the company’s values. However, research has told us that AI can be biased.  

However badly designed AI recruitment systems can widen inequality and increase labour market friction.  

Recent research led by Leiden University in the Netherlands, emphasises that AI can reproduce and perpetuate diversity bias, therefore discriminating against job applicants because of their personal characteristics. 

The researchers raise the example of the Amazon resume selection algorithm, with its design relying on historical gender imbalances in the company’s hiring practices, favouring men over women for technical roles.

They stress that while AI can support recruitment practices, it should not replace human judgment. Human oversight is essential to ensure fairness and accountability. 

Research from Pace University in the US finds that automated AI recruitment systems based on simplistic keyword screening can create what they call “artificial frictional unemployment” a scenario where job seekers remain unemployed for extended periods not because of a lack of available jobs or skills, but due to inefficiencies in automated hiring systems. This means capable candidates are excluded due to poor matching algorithms.  

Artificial frictional unemployment doesn’t just affect individuals. The researchers share that it suppresses productivity growth, slows wage progression, and risks creating a generation that never fully integrates into the labour market. 

To fix this, while still using AI in the recruitment process, the researchers suggest improving semantic matching between CVs and job requirements using skills-based matching instead of keyword matching. 

“The findings highlight the economic costs of outdated hiring systems and the potential gains from improving semantic alignment in labour market matching,” says the lead researcher, Ibrahim Denis Fofanah. 

The rise of skills-based hiring 

Isabel Berwick believes we will likely see a significant rise in skills-based hiring again this year.  

“We need to build a global workforce that can adapt, change, pivot all the active verbs to reflect the transformation that’s underway at work. And yet, most employers are still hiring on past experience and education, not skills and potential,” she says.

Companies are increasingly prioritising practical skills and experience over formal qualifications to address tight labour markets. The logic makes sense – with AI moving fast and economic instability on the rise, employers increasingly need people who can learn quickly, adapt and solve problems.

In response to this, a growing number of major companies, particularly in the technology, finance, and consulting sectors, have reduced or eliminated higher education degree requirements for many positions to focus on skills-based hiring.

For example, Walmart is focusing on skills over degrees, and made the move to skills-based hiring in 2023. Now, more than 90 percent of its US roles do not have a formal degree requirement. 

In response to this, many business schools and universities have also made a move towards skills-based learning, embedding experiential learning, live projects and industry partnerships into their curricula.  

Institutions such as Hult International Business School have built challenge-driven programmes around live business problems, aiming to narrow the gap between classroom and workplace. At Kingston Business School, a Future Skills framework helps students develop ‘human skills’ such as adaptability, collaboration and resilience alongside technical expertise. Meanwhile, schools including MIT Sloan School of Management have embedded action learning labs into their degrees, requiring students to work directly with organisations on strategic and operational challenges. 

The emphasis has changed in universities and business schools from what students know, to what they can do. This is very much a reflection of a changing job market, where human skills are much more in demand than ever before. 

‘Fix the house first before inviting people over’

However, even the very best hiring practices in the world will be incapable of building a perfect team if the company culture is toxic. Technology can optimise matching, but it cannot repair distrust, burnout or poor leadership.  

“We need to look inside and solve our own organisational issues, rather than just look for salvation in the form of brilliant new hires,” says Berwick. “I’d say fix up the house first and then invite some great new tenants.” 

Beyond protecting staff, increasing inclusivity and company culture is likely to have a positive impact on a company financially too. A study from Columbia University and Duke University found that 92% of CEOs and CFOs believe that improving their company’s culture would increase its value.

MIT Sloan highlighted three surprising ways organisations can promote a more positive work culture:

• People analytics: The researchers used natural language processing on more than 1.3 million Glassdoor reviews to capture how an employee rates their company This, they say, is a great way to better understand what matters to employees and how companies can change their work culture to be more positive

• Layoffs: Layoffs can sometimes have a positive impact on company culture. They can open up vacancies, giving leaders an opportunity to hire people who reflect the organisation’s culture, or the culture they want to aspire to, the research found.

• Use in-person time to its full potential: Organisations and leaders need to be more mindful about framing on-site time as an opportunity to shape culture and social norms

How to fix a broken job market

Yes, the job market is broken. But it’s broken because companies’ technological and organisational systems are not aligned with how work is evolving. Algorithms can’t make perfect decisions about people, and humanity is too messy and complicated to put in a singular box. AI also can’t replace the benefit of investing in real people.

To rebuild the job market, then, employers must invest in entry-level candidates and rebuild trust within organisations. Without that, the diamond will only grow sharper at the base. 

By, Chloë Lane

Enjoyed this article? You may like this…