'I applied for 200 jobs' - How 'spray and pray' AI job applications can backfire

When a 22-year-old geography graduate from Maidstone began his job hunt, he did what a growing number of his peers are doing: he turned to artificial intelligence. He used AI tools to help write cover letters, polish his CV and identify the keywords that automated screening systems look for. By the time he secured a job as a restaurant manager, he had applied to more than 200 positions — a process he described as exhausting and demoralising.

His experience captures a wider shift in the graduate labour market. AI has made it radically cheaper to apply for jobs, and candidates are responding by applying for far more of them. But new research suggests the strategy — which industry professionals call "spray and pray" — is beginning to backfire, as employers grow more adept at spotting AI-assisted applications and more sceptical of what they contain.

What the research shows

A survey conducted this summer by the Institute of Student Employers, which examined practices at 155 major UK graduate recruiters, paints a nuanced picture. Employers have, perhaps surprisingly, grown more receptive to candidates using AI at certain stages of the hiring process. Having AI help draft a CV or cover letter is now the most widely accepted use of the technology.

But acceptance has limits, and employers say candidates are crossing them. Fully 92% of the employers surveyed said they had encountered or suspected AI misconduct in recruitment — up from 85% the previous year. The most problematic behaviour, in recruiters' eyes, involves deploying AI as a real-time tool during video interviews or having it complete online assessments on a candidate's behalf. That is no longer a grey area: it is, employers say, cheating.

A young graduate at her laptop during a long job search, capturing the frustration of mass applications Credit: Jan Vašek / Wikimedia Commons, CC0

Separate data underlines how widespread the problem has become. A survey of 1,500 employers found that 34% of hiring managers had encountered job applications containing AI-generated falsehoods — not merely AI-polished prose, but fabricated qualifications and experiences. As one analysis put it, graduates are increasingly using AI to misrepresent themselves, and the line between legitimate enhancement and outright fraud is blurring.

The flood of applications

The clearest symptom of the spray-and-pray era is volume. Energy company EDF closed its graduate recruitment scheme just three weeks after opening it, having received 12,000 applications — a 10% increase on the previous year. Leadership at the company attributed the surge at least partly to AI-enabled mass application strategies.

EDF's experience is not unusual. When applying costs little more than a few prompts, rational candidates apply everywhere — and employers drown. The result is a kind of arms race: candidates use AI to get past the filters, employers deploy tougher filters to cope with the candidates, and the genuine signal of who is actually right for a role gets harder to detect on both sides.

That dynamic is precisely why spray and pray can backfire on the applicant. Applications fired off in bulk tend to be generic, and generic is exactly what both human recruiters and detection systems are learning to discount. A candidate who submits 200 AI-assisted applications may get fewer interviews than one who submits 20 carefully tailored ones — while also risking being flagged for the very AI assistance that made the volume possible.

How employers are responding

Employers are not standing still. EDF, for example, has restructured its process so that CVs are not reviewed until later stages of hiring, and now leans on skills-based and behaviour-focused assessments alongside video interviews specifically designed to detect scripted or AI-generated responses. The idea is to evaluate what candidates can actually do, rather than what their polished paperwork claims.

Even so, some candidates continue trying to circumvent the new measures — including, recruiters say, during interviews themselves. And the technological response is still in its early stages: only 16% of the employers surveyed currently use AI for candidate screening, though more than half indicate they may adopt such systems as application volumes keep rising.

Recruiters also stress a point that no detection system can fully automate: the qualities they are actually hiring for. Authenticity, personality and diverse thinking cannot be generated by a prompt, they say, and they remain the attributes that distinguish candidates once the paperwork is set aside. The Maidstone graduate's own account is telling in this respect: he used AI for cover letters, CVs and keywords, but drew a line at using it during assessments or interviews — a distinction many employers would recognise as the right one.

"We have no choice"

Not everyone accepts the employers' framing. Some graduates and labour advocates argue the problem is systemic rather than individual — that the hiring system itself creates the incentives it then punishes. When large employers filter thousands of applications through automated screening systems that scan for keywords before any human reads a word, candidates face a stark choice: optimise their materials for the algorithm, or risk being filtered out before they are ever seen.

From this perspective, graduates who embellish or AI-enhance their applications are responding rationally to a process that rewards optimisation. The sheer volume of applications for high-profile roles creates what labour advocates describe as a pressure cooker in which honesty is disadvantaged. Young candidates report feeling compelled to match their profiles to keyword-driven algorithms regardless of the ethical implications — a dynamic that suggests a fundamental mismatch between how companies recruit and how young professionals understand their own value in a saturated market.

There is also a generational fairness argument. Previous cohorts of graduates competed for jobs without algorithmic gatekeepers; today's graduates face screening systems, one-way video interviews and assessment centres designed to process them at industrial scale. If the process treats candidates as data to be filtered, some argue, candidates can hardly be blamed for treating it the same way.

Where this is heading

The most likely outcome is that both sides keep escalating. Employers will invest more in verification — the bnewso analysis of the trend notes that companies specialising in identity verification and workforce analytics have seen rising demand as corporations seek to detect inconsistencies in text and video interviews. Policymakers in the EU and the US are reviewing guidelines on algorithmic discrimination in hiring, and legal experts warn that a patchwork of international rules could complicate recruitment for global employers.

For graduates, the practical lesson from the research is narrower but important: AI is a tool, not a strategy. Using it to draft, polish and organise a job search is increasingly accepted and increasingly normal. Using it to fabricate experience, sit interviews by proxy or complete assessments is misconduct that employers are actively hunting — and being caught can end a candidacy instantly. The candidates who thrive will be those who use AI to present themselves more clearly, not those who use it to become someone else.

The deeper question is whether the hiring system itself will change. As long as employers filter by keyword and candidates optimise by keyword, the spray-and-pray equilibrium holds: more applications, more AI, more suspicion, more filtering. Breaking it would require hiring processes that reward genuine fit over volume — skills-based assessment, human review earlier in the funnel, and a recognition that the best candidate is not always the best-optimised one. Until then, graduates will keep firing off applications by the hundred, and employers will keep wondering which of them are real.

*Reporting for this story drew on RocketNews, BNewsO, NewsLocker's ICT desk, NewsLocker's project management desk and NewsLocker's South East regional roundup. For more on work and the economy, see our Business and Technology desks, our report on the September US jobs slowdown, the Bradford teens' social media trial on young people's digital lives, 'Don't call us snowflakes' on how young people see themselves, and our World and News coverage.*