Nearly three quarters of students and graduates are using AI during job applications, even as many remain uncomfortable with employers using the same technology to assess them.
New research cited by People Management and published by the Institute of Student Employers found that 73% of respondents had used AI at some point during the application process. That is up from 55% last year, showing how quickly generative AI has become embedded in early-career job search behaviour.
The most common use was writing CVs and cover letters, with 46% of respondents saying they had used AI for that purpose. Yet only 38% thought it was acceptable for employers to use AI to screen CVs.
The findings show a widening fairness gap in graduate recruitment. Candidates are increasingly comfortable using AI to improve applications, structure answers, and compete in high-volume recruitment markets. At the same time, they are wary of employers using AI to make evaluative judgements, particularly in online interviews and virtual assessments.
Qualification screening was the only area where AI use by employers was seen as broadly acceptable. Candidates appeared more comfortable with AI supporting administration or logistics than with AI judging suitability, communication, personality, motivation, or potential.
Graduate recruitment is becoming more competitive. Separate findings from Prospects at Jisc’s Early Careers Survey of 5,000 students and graduates found that high-volume applications are rising, with respondents submitting more than 100 applications increasing from 8% to 13% year on year. Recent graduates faced the greatest pressure, with 20% saying they had applied for more than 100 roles.
The spread of AI tools may intensify that volume problem. If candidates can generate tailored applications more quickly, employers may receive more submissions while finding it harder to distinguish genuine fit from polished generic material. Recruiters may then turn to automated screening, creating a feedback loop in which applicants and employers both rely more heavily on AI.
That feedback loop is already reshaping early-career hiring. Employers have to decide whether AI-assisted applications represent reasonable tool use, misrepresentation, or a new baseline skill. Candidates, meanwhile, face inconsistent rules across sectors. Some employers allow AI use if disclosed, others discourage it, and many provide little clear guidance.
Students entering the labour market are already building AI into study, collaboration, and commercial activity. Recent Gen Z research showed AI and digital tools reshaping how young people learn, create, and build online ventures, which means graduate employers are recruiting from a cohort for whom AI-assisted work is increasingly normal rather than exceptional.
The challenge is not simply whether candidates use AI, but what that use reveals. A well-structured AI-assisted CV may say less about a candidate’s writing ability than a traditional application, but it may say more about their ability to prompt, refine, edit, and use tools effectively. Employers need to decide which capabilities they are actually assessing.
Traditional application stages are under pressure as a result. CVs, cover letters, competency statements, and online written questions were already imperfect signals of potential. Generative AI weakens them further if employers treat them as pure evidence of unaided skill. Assessment design may need to move closer to supervised tasks, work simulations, structured interviews, portfolio review, or exercises that test judgement and reasoning in real time.
There is also a diversity and access dimension. AI tools can help candidates who lack professional networks, coaching, or confidence to present their experience more clearly. They may also favour those who know how to use paid tools well or who understand the language of corporate recruitment. Blanket bans could be difficult to enforce and may penalise honest applicants while doing little to stop undisclosed use.
Employer use of AI requires its own governance. Automated screening tools can process large applicant volumes, but they must be tested for bias, explainability, accuracy, and proportionality. Candidate resistance to AI assessment suggests that transparency will become part of employer brand. Organisations that use AI without explaining its role may face lower trust, particularly among early-career applicants who already feel the market is difficult to navigate.
The trend also lands amid concern that AI is narrowing entry-level pathways. Major professional services companies have already cut graduate roles as automation changes junior work, raising questions over how companies will train the next generation if lower-level tasks are increasingly automated.
A clearer recruitment compact is now needed. Candidates should be told what AI use is acceptable, what must be disclosed, and which parts of the process must be completed unaided. Employers should also explain where they use AI, what decisions humans retain, and how assessments are checked for fairness.
AI is now part of the graduate labour market. Treating it as an exception will become increasingly unrealistic. Recruitment processes will need to measure capability, integrity, and potential in an environment where both applicants and recruiters have access to powerful tools.
You must be logged in to post a comment.