How to Use ChatGPT to Prepare for an Interview: 8 Example Prompts

A few years ago, practicing for an interview meant rehearsing in front of a mirror, asking a patient friend to play the interviewer, or combing through Glassdoor question dumps hoping you'd guess right. Today, AI tools like ChatGPT, Claude, and Gemini have changed the equation entirely. You now have access to a tireless, infinitely patient practice partner available at 2 a.m. the night before your big interview — one that can roleplay a hostile interviewer, critique your STAR stories with surgical precision, and generate 30 role-specific technical questions in under ten seconds.

This guide is not about handing your job search over to a chatbot. It's about using AI strategically so that you show up sharper, more confident, and better prepared than the person interviewing before you.

Why AI Has Become a Genuine Game-Changer for Interview Prep

The traditional approach to interview preparation has a fundamental flaw: feedback loops are slow and expensive. Career coaches charge hundreds of dollars per session. Friends get tired of asking the same behavioral questions. Mock interview platforms offer limited flexibility. And none of these options are available at midnight when nerves kick in.

AI flips all of that. The advantages are practical, not theoretical:

Always available. You can run a full mock behavioral interview at any hour, as many times as you need, without scheduling anyone.

Endlessly patient. You can give a rambling, 400-word answer to "Tell me about yourself" and ask for feedback without embarrassment. AI won't get bored or sugarcoat a weak response the way a friend might.

Genuinely personalized. Feed the AI your resume, the job description, and the company name, and it will generate questions specific to your background and this role — not generic questions pulled from a listicle.

A safe environment to fail. Bombing an answer in practice costs nothing. The more you fail safely in prep, the less you fail when it counts.

Setting Up AI for Success: Give It the Right Context

The single biggest mistake candidates make when using AI for interview prep is treating it like a search engine — typing "give me interview questions" and expecting magic. The quality of AI output scales directly with the quality of the context you provide.

Before you start any mock interview session, paste the following into a single opening message:

1. The job description — the full text, not just the title. AI will extract the keywords, required skills, and implied competencies the interviewer will probe.
2. Your resume summary or bullet points — so the AI can ask questions grounded in your experience, not generic ones.
3. The company name and any recent news — a quick paste from their About page or a recent press release lets the AI simulate culturally relevant questions.

An effective setup message looks like this:

"I have an interview next week for a Senior Product Manager role at Stripe. Here is the job description: [paste]. Here is a summary of my background: [paste your resume highlights]. Stripe recently launched their new embedded finance product. Use all of this context for all of our practice today."

From this point forward, every prompt you send will be interpreted against that context. Do not skip this step.

The 8 Essential Prompts

1. Act as the Interviewer (Behavioral Round)

"Act as a hiring manager for this role. Conduct a 5-question behavioral interview with me. Ask questions one at a time and wait for my full answer before asking the next. After all 5 answers, give me a detailed assessment of each response, noting what worked and what could be stronger."

This prompt forces a realistic interview cadence. The AI won't dump all five questions at once — it will wait for you, just like a real interviewer would. The post-session assessment is where the real value lies: you get written feedback on every answer in one place, which you can review and act on.

2. Improve My Answer / Make It More Concise

"Here is my answer to the question 'Tell me about a time you handled a conflict with a colleague.' Score it from 1–10 on clarity, structure (STAR method), and relevance to the role. Then rewrite it as an ideal 90-second answer: [paste your answer]."

Verbosity is one of the most common interview killers. This prompt is blunt: it scores you, then shows you what a tight version of your own story looks like. Use the rewrite as a benchmark, not a script — then try to match its energy in your own words.

3. Generate Role-Specific Technical Questions

"Generate 12 technical interview questions for this role, ranging from foundational to advanced. For each question, note what the interviewer is really trying to assess."

Generic technical question lists are useless. This prompt draws from the specific technologies and competencies mentioned in the job description you already provided. The "what they're really assessing" annotation is invaluable — it tells you whether a question is testing depth of knowledge, debugging instinct, system design thinking, or something else entirely.

4. Mock Salary Negotiation Roleplay

"Let's roleplay a salary negotiation. You are the recruiter. Open by telling me the offer is £68,000 base — which is £7,000 below my target of £75,000. I will negotiate. Push back realistically two or three times before accepting a reasonable counter. After we finish, tell me what I did well and what I should have said differently."

Most candidates never practice negotiation out loud. Doing it in text with an AI that pushes back realistically is remarkably effective. The debrief at the end surfaces hesitations in your language ("I was hoping for..." versus "I'm targeting...") that you'd never notice otherwise.

5. Help Me Research This Company

"I have an interview with [Company]. Here is their About page: [paste]. Here are three recent news items about them: [paste]. Based on this, what are the company's current strategic priorities? What challenges might they be facing? Generate 5 questions I could ask that would signal I've done serious research."

AI can't browse the internet in real time (in most configurations), so you need to feed it the raw material. But once you do, it synthesizes information faster than you ever could manually. The output gives you informed talking points and questions that go well beyond "I read your website."

6. Critique My STAR Story for a Specific Competency

"I am going to tell you a STAR story for the competency 'stakeholder management.' After I share it, critique it on three dimensions: (1) Is the Situation specific and believable? (2) Does the Action section focus on what I did vs. what 'we' did? (3) Is the Result quantified and memorable? Here is my story: [paste]."

The STAR method is well-known but rarely executed well. The most common failure is an Action section full of "we" — which obscures your personal contribution. This prompt specifically hunts for that. It also pushes you to quantify your result, which most candidates skip.

7. Generate Questions I Should Ask the Interviewer

"Based on this job description and what you know about this stage of the interview process (final round with the hiring manager), generate 8 questions I could ask. Flag which 3 are most likely to impress a senior hiring manager and why."

"Do you have any questions for us?" is never filler. Candidates who ask sharp, specific questions signal genuine interest and strategic thinking. This prompt generates a ranked list with reasoning — so you're not just picking questions randomly, you're choosing them strategically.

8. Create a Study Guide from the Job Description

"Read this job description carefully. Create a structured study guide I can use in the 48 hours before my interview. Include: key technical areas to review, likely behavioral themes based on the role, a list of my probable weak spots given my background, and a 60-minute mock session plan."

This prompt is best used two or three days before the interview. It turns the job description into a personalized revision plan — one that identifies your specific gaps rather than giving you a generic checklist. The 60-minute session plan structures your time so you don't spend four hours on things you already know.

The AI-Prep Loop: A Simple Workflow

Effective AI-assisted prep is not a one-and-done exercise. It's a loop. Here's how the cycle works in practice:

Each loop tightens your answers. The goal is to run this cycle at least three times for the questions you're least confident about before the actual interview.

The Authenticity Warning: Why Memorizing AI Answers Is a Trap

This is important enough to say plainly: AI-generated answers will make you sound like AI.

Interviewers who conduct many interviews in a week will notice something is off — responses that are perfectly structured but feel hollow, stories that are technically correct but lack the specific texture of lived experience. The giveaway is usually the absence of small details: the name of the client you were dealing with, the exact emotion you felt when the project failed, the specific reason you made the call you made.

AI can structure your ideas. It cannot supply the raw material. Use it to identify weak spots in your reasoning, tighten your phrasing, and eliminate filler — but always feed your own genuine memories and experience into the process. The best way to do this:

1. Draft your answer in your own words first, without looking at any AI output.
2. Paste your answer into the AI and ask for a critique.
3. Revise based on the critique — in your own words again.
4. Practice saying the revised answer out loud, not re-reading it.

The final answer should sound like a sharper version of you, not a different person entirely.

Know the Limitations

AI is not a replacement for real interview practice, and it has specific blind spots you should be aware of:

It doesn't know the company's internal culture. AI can tell you what a company says about its culture on its website. It cannot tell you that the hiring manager tends to prioritize speed over thoroughness, or that the team has high turnover, or that the culture description doesn't match reality. For that, you need people — LinkedIn connections, Glassdoor reviews, people who have worked there.

It can hallucinate facts about a company. If you ask AI to tell you about a company's recent product launches or financial performance without providing source material, it may confidently state things that are wrong. Always verify any company-specific facts AI produces against primary sources before using them in an interview.

It can't replicate a real interviewer's intuition. AI will rarely ask a follow-up question designed to probe whether you're exaggerating a claim. Real interviewers do this constantly. After you give an answer, ask the AI explicitly: "What follow-up question would a skeptical interviewer ask me right now?" This partially compensates for the limitation.

Tone and delivery don't come through in text. AI can tell you that your answer is well-structured, but it cannot tell you that you're speaking too fast, looking at the floor, or using too many filler words. For that, record yourself on video or audio, even just on your phone.

Structuring a 60-Minute AI Mock Interview Session

Here is a practical session structure you can follow in the 48 hours before your interview:

Minutes 0–10: Context setup. Paste the job description, your resume summary, and any company research into the chat. Ask the AI to confirm it understands the role and your background before proceeding.

Minutes 10–30: Behavioral round. Ask the AI to run a 5-question behavioral interview, one question at a time. Answer each one out loud (or type at the pace you'd speak), then move on. Don't stop to self-correct mid-answer — simulate real conditions.

Minutes 30–40: Feedback and revision. Ask for a full assessment of your five answers. Identify the two weakest answers and rerun them once each.

Minutes 40–52: Technical or role-specific questions. Ask for 6 technical questions at increasing difficulty. Answer each briefly (2–3 minutes each in a real interview). Ask for corrections where you were wrong.

Minutes 52–58: Questions to ask the interviewer. Generate and review your shortlist. Choose 3 that feel authentic and that you could discuss further if the interviewer asks you to elaborate.

Minutes 58–60: Wind down. Ask the AI: "What is the single most important thing I should remember to do differently tomorrow?" Often the answer is obvious once it's stated plainly.

Advanced Tips for Getting More Out of AI Prep

Ask it to get harder as you improve. After a few sessions, tell the AI: "My answers are getting more polished. Start asking more probing follow-up questions and push back harder on claims I make." This simulates late-stage interviews, which are often more rigorous.

Practice edge cases. Most interview prep focuses on standard questions. But edge cases are where candidates most often get caught flat-footed. Use prompts like: "Ask me about the 18-month gap on my resume and push back on my explanation." Or: "Ask me an unusual question designed to test how I think under pressure." Or: "Play an interviewer who seems skeptical about my ability to do this job and keep that skepticism throughout."

Use it for specific competency gaps. If you know a role requires stakeholder management skills and that's not your strongest suit, run an entire focused session on that one competency — five or six different behavioral questions, all targeting the same theme, until you have a library of genuine stories that demonstrate it.

Cross-reference across AI tools. Different AI systems have different strengths. ChatGPT tends to be thorough and structured. Claude tends to be more nuanced in assessing the emotional texture of answers. Gemini integrates well with recent web data when available. There is no rule that says you have to use only one.

The bottom line is this: the candidates who use AI well in their interview prep are not the ones asking it to write their answers. They're the ones using it to stress-test their own thinking, identify their own weak spots, and practice the uncomfortable parts of an interview over and over until they aren't uncomfortable anymore. That's the edge AI actually gives you — and it's a real one.