AI in Job Applications: What UK Employers Accept
A new survey of large UK graduate employers shows record levels of suspected cheating, mostly through undeclared AI in interviews. Here is what it says, where employers draw the line, and how to use AI in your applications without putting your credibility at risk.
What the 2026 survey found
On 14 October 2026, coverage of the Institute of Student Employers (ISE) annual student recruitment survey put a number on something recruiters have been grumbling about for a while. The survey covers 155 large employers. Of those, 92 per cent had come across possible or verified cheating at some point in their recruitment process, and 66 per cent had flagged the use of unapproved AI tools in interviews.[1]
The trend is not new, it is accelerating. Another report on the same survey says the 92 per cent compares with 85 per cent a year earlier and 78 per cent the year before that. The most common form is a graduate using AI in an interview without saying so. Online assessments come next.[2]
Two details matter if you are applying right now:
- The problem is concentrated in live and timed stages. Interviews and online assessments are where employers report most of their concerns. CVs and cover letters come far lower in the same report.[2]
- Few employers ban AI outright. The survey summary says most want to set clear limits on appropriate use rather than forbid it.[1]
In other words, using AI is not automatically a problem. Using it in the wrong place, or without telling anyone when you were asked to, is.
Where employers draw the line
The reports do not publish a single rulebook, and each employer sets its own. But the pattern in what recruiters describe is consistent: the closer a task is to proving what you can do on the day, the less tolerance there is for outside help.
| Stage | How employers tend to see AI help |
|---|---|
| Researching the company and role | Generally fine: it is preparation |
| Improving the wording of your CV or letter | Often accepted, if every fact is true and yours |
| Written application questions | Check the instructions: some employers state limits |
| Online tests and timed exercises | High risk: designed to show your own work |
| Live interviews (video or phone) | Highest risk: undeclared real-time help is what employers report most |
This table is our reading of the reports rather than an official standard. When in doubt, the employer's own instructions win.
How recruiters say they spot it
A Manchester recruiter quoted in the coverage says answers produced with AI help are easy to spot. He describes candidates who seem to be reading off an autocue, with delays before answering, filler phrases and unusually formal language. He also says candidates who use real-time answer apps in virtual interviews are dropped from the process.[2]
The ISE chief executive, Stephen Isherwood, makes a point that goes beyond detection: if you misrepresent yourself in an interview, you could end up in a job that does not suit you.[2] That is a practical argument, not just an ethical one. An offer won with borrowed answers still leaves you doing the job with your own skills.
A safe way to use AI in your applications
Nothing in these reports says you should avoid AI when preparing. What they suggest is a simple split between preparing and performing.
- Read the instructions first. Application forms and assessment invitations increasingly say what is allowed. If a stage tells you not to use AI tools, do not.
- Start from your real experience. Give any writing help your actual projects, results and dates. Never let a tool add a skill, a grade or a responsibility you cannot defend in a conversation.
- Read every line out loud. If a sentence in your CV or answer sounds like someone else, rewrite it in your own words. You will be asked about it.
- Prepare examples, not scripts. Before an interview, write down a few real situations with what you did and what happened. This is what you can talk about with no assistance, and it is far harder to fake than a memorised paragraph.
- Keep tools out of the live stage. If you want to practise, practise beforehand, then close everything and talk.
- Be ready to say how you used AI. If an application asks, answer plainly. Saying "I used a tool to tighten the wording of my CV, the content is mine" is much safer than being found out.
What this means for graduates in a tough market
The same coverage describes a market that is recovering only slowly: graduate hiring fell by 8 per cent last year and rose by 4 per cent in 2026, with employers expecting a further 7 per cent rise over the next 12 months.[2] Placement and internship opportunities also fell by 8 per cent.[1] More competition for fewer places is exactly the situation in which shortcuts feel tempting.
The better response is to make your honest material stronger. A CV tailored to the specific posting, with real evidence for each requirement, costs more effort than a generic one, but it holds up when somebody asks you about it. Our guide on tailoring your CV to a job posting explains how, and the interview preparation guide covers how to build answers from your own examples.
Frequently asked questions
Is it cheating to use AI to write my CV or cover letter?
The reports describe CVs and cover letters as a much smaller part of the problem than interviews and online assessments. Many employers accept help with wording. What they object to is misrepresentation: skills or achievements that are not yours. Check the employer's instructions and keep every claim true.
Can I use AI to practise for an interview?
Practising beforehand is different from getting live help during the interview itself. The concern in the survey is the use of unapproved tools during interviews. Practice questions, mock answers and feedback before the day are preparation.
What should I do if an application says AI is not allowed?
Follow it. Write that stage yourself, from your own experience, and keep it short and specific. A plain, accurate answer is better than a polished one you cannot explain.
Do these findings apply outside the UK?
The survey covers large UK graduate employers, so the figures describe that market only. Practices and rules differ elsewhere, so check local guidance before assuming the same limits apply.