As of 13 August 2026, AI cannot detect cheating in a remote job interview.

What to actually do

1. Hand it to a person

   The route this page recommends

   A person who owns the outcome does this end to end, worth it when the failure is dear.

2. Use a tool built for this

   Second choice

3. Do it yourself

   The distant third

   A chat interface gets you a draft, but you cannot verify it yourself. That is the catch.

   How to actually do it

   1. Open the interview invitation, candidate instructions and assessment policy, then write down exactly what outside help and materials were permitted.
   2. Confirm that the recording, transcript or monitoring data was collected with the required notice and consent, and remove personal data that is not needed for the review.
   3. Gather the interview questions, candidate answers, time-stamped notes, agreed reasonable adjustments and any permitted reference materials in one document.
   4. Paste that material into a chatbot with the supplied prompt and ask it to list observable events without calling any event cheating.
   5. Compare each flagged event with the original recording and interview notes, then record an alternative explanation or mark the point as unverified.
   6. Ask a trained interviewer or HR adviser to run the neutral follow-up or repeat assessment under the same rules, and keep the final decision separate from the model's analysis.

   Prompt

   You are helping an employer review a remote job interview for possible breaches of the interview rules. Do not decide whether the candidate cheated and do not infer dishonesty from eye movement, facial expression, accent, disability, neurodivergence, pauses or other personal characteristics. Use only the evidence I provide.
   
   Interview rules:
   [PASTE THE RULES THE CANDIDATE WAS GIVEN]
   
   Interview transcript or time-stamped notes:
   [PASTE THE TRANSCRIPT OR NOTES]
   
   Permitted materials and reasonable adjustments:
   [PASTE THESE, OR WRITE NONE KNOWN]
   
   Produce four sections:
   1. Observable events, with exact quotations or timestamps where available.
   2. For each event, possible explanations that do not involve cheating and possible explanations that do, without treating either as established fact.
   3. Evidence gaps and what cannot be concluded from this material.
   4. Neutral, job-relevant follow-up questions or a repeat assessment that applies the same process to candidates consistently.
   
   Do not assign a probability, score or verdict. Do not recommend rejection. Mark any point that would require a human interviewer, HR adviser or employment solicitor to assess. Treat the material as confidential and do not ask for unnecessary personal data.

   Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.