As of 13 August 2026, AI can create interview questions from a job description.
This still needs a person who signs their name to it.
Can you do it?
5 minutesto a draft.
30 minutesto something you’d act on.
Cost, all in£0
Skill neededchat-fluent
Who has to check ita colleague
What the alternative costsA human hiring manager or HR adviser remains the alternative; no price is stated here.
If this goes wrong: the questions assess irrelevant or potentially discriminatory factors, producing an unfair interview and a poor hiring decision.
What to actually do
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.
Use a tool built for this
Second choiceDo it yourself
The distant thirdA chat interface, chat-fluent skill, and roughly 30 minutes until you can act on the result.
How to actually do it
- Open the final job description and remove outdated duties, unnecessary requirements and any personal information before copying it.
- Gather the essential criteria, desirable criteria, competencies, interview length, interviewers and interview format from the hiring manager or recruitment brief.
- Paste the job description and those details into a chatbot using the supplied prompt, then ask it to produce the question set and scoring guide.
- Open the output beside the job description and compare every question with the criterion it is meant to assess, deleting questions that test irrelevant personal preferences or invented requirements.
- Ask an HR colleague to check the questions, follow-ups, scoring guide, protected-characteristic exclusions and reasonable-adjustment note before the interview is scheduled.
- Send the approved question set and scoring guide to all interviewers, and agree that they will record evidence against the same criteria for every candidate.
Prompt
Create a structured interview question set for the UK job described below. Job description: [PASTE THE FULL JOB DESCRIPTION] Interview details: - Interview stage: [for example, first interview or final interview] - Length: [NUMBER OF MINUTES] - Interviewers: [ROLES OR NAMES] - Format: [for example, panel, video or in person] - Number of questions wanted: [NUMBER] Assessment requirements: - Essential criteria: [PASTE THEM] - Desirable criteria: [PASTE THEM] - Behaviours or competencies to assess: [PASTE THEM] - Any practical exercise or role-specific evidence required: [PASTE DETAILS OR WRITE NONE] Produce: 1. A brief assessment plan showing which criterion each question tests. 2. A balanced set of open questions, with no question testing more than one main criterion unless necessary. 3. Two neutral follow-up prompts for each main question. 4. A simple scoring guide from 1 to 5 with observable evidence for each score. 5. A short list of acceptable evidence and red flags for each question. Red flags must concern evidence relevant to the role, not personality or protected characteristics. 6. A suggested order that fits the stated interview length. 7. Two questions the candidate can ask at the end. Use plain UK English. Base questions only on the information supplied. Do not invent duties, qualifications or company facts. Do not ask about age, disability, health, pregnancy, family plans, religion, race, sex, sexual orientation, gender reassignment, marital status or other protected characteristics. Do not infer anything about a candidate from their name, accent, appearance or background. Include a note telling the interviewers to apply the same core questions and scoring criteria to every candidate, while allowing reasonable adjustments. Flag any requirement in the job description that could create a fairness or accessibility concern for an HR reviewer rather than deciding the legal position yourself.
Open it prefilled in ChatGPT or Claude, or copy it into Gemini, which takes no prefill link.
What it gets wrong
- AI cannot know which parts of a vague job description reflect the real priorities of the hiring manager.
- AI cannot decide whether an essential requirement is genuinely necessary, proportionate or accessible for the role.
- AI cannot take responsibility for consistent interviewing, reasonable adjustments or the employment decision.
- AI cannot reliably spot every subtle source of bias in the wording or scoring approach without informed human review.
Even on a YES, the friction has a name: judgement under ambiguity, legal accountability and verification cost.
How we scored this
Five axes, each scored nought to two by hand: ten means AI carries the task cleanly, and the thresholds that turn a total into YES, PARTLY or NO are published in the methodology. Each axis name links to its definition.
| Axis | Score (0–2) |
|---|---|
| Output | 2 |
| Inputs | 2 |
| Verification | 1 |
| Liability | 1 |
| Effort delta | 2 |
| Total | 8 / 10 |
The methodology and its thresholds are published in full.
FAQ
- Can ChatGPT create interview questions from a job description?
- Yes. It can produce relevant questions, follow-ups and a basic scoring guide from a clear job description, but a hiring manager or HR colleague should approve the result before use.
- Are AI-generated interview questions fair?
- Not automatically. AI may repeat vague requirements or create questions that disadvantage some candidates, so check every question against the role and use the same core questions and scoring criteria for everyone.
- Can AI create competency-based interview questions?
- Yes. Give it the competencies, behaviours and evidence you need to assess, and ask for structured questions with observable scoring criteria. It cannot decide whether those competencies are the right ones for your organisation.
- Do I need HR to check AI interview questions?
- For a routine role, a knowledgeable hiring colleague can check relevance, consistency and clarity, but HR review is sensible where the role or selection process creates legal or accessibility concerns. Your organisation remains responsible for the fairness of the interview.
Nearby answers
Assessed by gpt-5.6-luna (gpt-5.6-luna) on 2026-08-13, second-checked by an independent model. Wrong somewhere? Email [email protected] and it gets re-checked.
The newsletter
AI news, new answers and product picks, straight to your inbox.