Can UK Employers Tell If I Used AI in My Application or Interview?

Key facts

The specifics, before the explanation.

• ISE survey, October 2026: 92% of UK graduate employers reported candidate misconduct and 66% saw AI misuse in interviews
• Deloitte UK reinstated in-person interviews for graduate roles from September 2026 in direct response
• Common detection: eye movement away from camera, pauses before fluent answers, identical phrasing across candidates, and probing follow-ups
• Most employers publish an AI policy in the application process; many ask you to declare how you used AI
• Under UK law an offer obtained by misrepresentation can be withdrawn, and professional bodies can treat it as a fitness issue

What this actually means

The scale of the problem is what has changed in 2026. Graduate employers used to see occasional copied cover letters. Now, according to the ISE, two thirds have seen candidates reading AI-generated answers during interviews, using hidden earpieces or a second screen, or submitting assessment answers that were clearly machine-written. Employers have responded by moving final stages back in person, adding proctoring to online tests, and training interviewers to probe.

Detection is less technical than people assume. Interviewers notice eyes tracking across a screen, a pause before a suspiciously polished answer, vocabulary that does not match the candidate’s CV or earlier answers, and the inability to go one level deeper when asked a follow-up. Some platforms flag tab switches, multiple faces or audio from a second device. Written answers are compared across the applicant pool for identical structures and phrases.

The rules are usually set out in the employer’s own policy, and most now have one. The typical line is that you may use AI to research the company, understand the role and practise, but you may not use it to generate answers during a live or recorded interview, and you must declare its use if asked. Breaching that policy is treated as misconduct, and because an offer rests on honest representations, it can be withdrawn even after you have accepted.

The safe route is to use AI the way employers expect: to prepare. Generate practice questions, rehearse your own answers out loud, review a transcript to see where you lost the thread, and then walk into the interview without any device helping you. That is also the preparation that survives a follow-up question.

What to actually do

Rules and figures move. Confirm anything time-sensitive with the official regulator before you rely on it.

• Read the employer’s AI policy before applying and answer any declaration question honestly
• Use AI to generate practice questions, then rehearse your own answers out loud and review the transcript
• Never have a second device or screen running during a live or recorded interview
• Make sure every example in your written application is one you can expand on for five minutes without notes
• If you used AI to tidy your cover letter wording, be ready to say so and to explain the substance in your own words

Frequently asked questions

Is it cheating to use ChatGPT to prepare for an interview?

No. Almost every UK employer allows AI for research and practice. The line is crossed when AI produces the words you submit or say during the assessment itself.

Can an offer be withdrawn if AI misuse is found later?

Yes. Offers are made on the basis of your representations, and many graduate contracts allow withdrawal for misconduct in the selection process. Regulated professions may also report it.

Will employers tell me if they suspect AI misuse?

Not always. Many simply reject the application. Some will ask follow-up questions or invite you to an in-person stage to check. If asked directly, answer honestly.

Sources

Referenced for this guide:

• Institute of Student Employers