AI is being used in the recruitment market both by the employers and the potential employees, but it is not always used effectively. Candidates are using AI to apply for a huge variety of jobs which do not always match their skillsets and HR departments are using it to filter through applications, sometimes rejecting someone who would suit the role well. Louis Allain, VP Product at Welcome to the Jungle, a media and recruitment technology company, discusses how companies may use AI effectively in the hiring process.
There’s a conversation happening in boardrooms and HR departments across the UK right now, along the lines of: ‘we’re receiving more applications than ever, our hiring process is more automated than ever, and yet we’re struggling more than ever to find the right people’. How is that possible?
The answer is that the UK jobs market has engineered itself into a vicious cycle, and most of the organisations caught in it don’t realise they’re part of what’s sustaining it.
More applications, worse outcomes
Welcome to the Jungle’s own data tells the story clearly: job applications on the platform grew by 28.1% year-on-year in the first quarter of 2026, and are now more than 50% higher than two years ago. That sounds like a healthy talent pipeline until you look at the quality: 77% of those applications don’t match the role they’re applying for. The sheer volume of candidates landing in hiring managers’ inboxes has grown dramatically, but the proportion worth progressing has not kept pace.
This isn’t a coincidence. It’s the predictable result of AI tools on the candidate side dramatically lowering the cost and effort of applying. When generating a tailored cover letter takes just 30 seconds and an application can be submitted with one click, people apply more broadly, more speculatively and with less consideration of fit. The friction that once served as a natural filter has been engineered away.
The automation trap
The logical response from employers has been to automate the other side of the equation too. If candidates are using AI to generate applications at scale, use AI to screen them at scale too. Deploy an ATS, score CVs automatically, set keyword filters and accept or decline at volume.
This is where the cycle tightens. Automated screening tools, particularly those trained on historical hiring data, tend to replicate the patterns of past decisions. They learn what previous hires looked like and filter for more of the same.
At the same time, candidates quickly learn how automated screening works, and they optimise for it. More keyword-matching, more CV padding and more applications per person, which generates more volume for employers to manage. This, in turn, again prompts more automation.
What gets lost in this exchange is exactly what both sides are actually looking for: a genuine match. Not a CV that clears a filter, but a person who will thrive in the role and the organisation.
The cost to business
For CXOs, the implications of this cycle extend well beyond recruitment. Poor hiring decisions are expensive – the cost of a mis-hire at a senior level is routinely estimated at multiple times annual salary when you factor in lost productivity, team disruption and the process of replacing them. But the subtler cost is what the cycle does to talent strategy over time.
When screening is automated and optimised for pattern-matching, organisations end up narrowing their own talent pools. They hire people who look like the people they’ve always hired. The unconventional thinkers and candidates who bring genuinely different experience are precisely the people most likely to be filtered out. This is a strategic issue dressed up as an operational one.
There’s also the question of employer brand. Candidates who experience a hiring process that feels opaque, automated and impersonal don’t forget it. In a market where employer reputation increasingly shapes who you’re able to attract, a poor candidate experience is a slow leak in the pipeline.
What breaking the cycle requires
The answer isn’t to reject AI in hiring – that ship has sailed, and used well, AI genuinely improves recruitment. The answer is to be deliberate about what AI should, and shouldn’t, be doing.
AI earns its place in recruitment when it removes administrative friction: drafting job descriptions, pre-screening against specific and transparent criteria, reducing the time recruiters spend on scheduling and co-ordination. Our own data suggests this kind of targeted automation can reduce shortlisting time from around three hours to 10 minutes per role. That’s real capacity returned to hiring managers.
What AI should not be doing is making decisions about people, or creating the conditions in which those decisions happen invisibly, without human oversight. The final call on whether a candidate progresses or is rejected should remain with a person. Not as a rubber stamp on an algorithmic output, but as a genuine exercise of judgement informed by everything the process has surfaced.
That distinction matters more than it might appear and is the difference between AI that expands what’s possible in hiring and AI that merely accelerates a broken process.
Measuring what matters
The good news is that the cycle isn’t inevitable. However, breaking it does require organisations to measure different things, and to be more deliberate about what they’re optimising for. The incentives that have sustained it – prioritising application volume, automating for speed, measuring throughput rather than match quality – have often developed gradually rather than by design.
Shifting that means tracking different outcomes: not how many CVs were processed, but how many hires were still in the role at 12 months. Not how quickly a vacancy was filled, but whether the person in it belongs there.
That’s a harder discipline than deploying another tool. But it’s the only way to stop the cycle from compounding – and to start using AI in recruitment for what it’s actually capable of: helping organisations find the people they’re genuinely looking for.
