- Why the generic prompt gives you a generic answer
- Prompts to research the role and company before you write anything
- Prompts to build STAR stories from your actual history
- Prompts for the tough, off-script questions
- Prompts for salary, offer, and negotiation questions
- The uncomfortable part nobody wants to say out loud
- A step-by-step 20-minute prep session
- What to avoid entirely
- Frequently asked questions
- Primary sources
The short version: good ChatGPT prompts for job interviews are specific to the job posting, your own history, and the format of the interview, not generic "tell me about yourself" filler you can find on any listicle. Feed the tool the actual job description, your CV, and the company's own words, and it will build you tailored questions, STAR stories, and follow-up answers in about twenty minutes. Copy the exact wording, but never the answers word for word, because the interviewer has almost certainly heard the same AI-generated answer from three other candidates this week.
Why the generic prompt gives you a generic answer
I've watched this happen more than once with people I coach. Someone types "give me interview questions for a marketing manager role" into ChatGPT, gets back the same twelve questions every career blog has published since 2019, rehearses them, and walks into the room sounding like a LinkedIn carousel. The problem isn't ChatGPT. The problem is the input.
Last year I sat with a client, I'll call her Sarah, before her second-round interview for a marketing director role at a mid-size fintech in Manchester. She'd used ChatGPT the night before with a bare prompt like "help me prepare for a job interview" and came out with an answer to "what's your biggest weakness" that sounded like it had been written by a committee. Polished, safe, and forgettable. We rebuilt the prompt together, pasted in the actual job listing plus the company's last two press releases, and asked ChatGPT to identify the three skills the posting mentioned twice or more. Turns out "regulatory compliance in financial marketing" showed up in the listing three times but Sarah's original answers never touched it. That one fix changed her prep completely, and she got the offer nine days later.
That's the difference specificity makes. A prompt built from the actual job description beats a generic prompt every time, because the interviewer isn't testing whether you can answer a question, they're testing whether you understood their specific problem.
Prompts to research the role and company before you write anything
Do this first, before touching a single practice question. Paste the full job description into ChatGPT and try:
- "Read this job description [paste it]. List the five skills or requirements mentioned most often, ranked by frequency."
- "Based on this job description, what problem is this company likely trying to solve by hiring this role right now?"
- "Summarise this company's last three press releases or blog posts [paste links or text] in plain language, and tell me what themes keep repeating."
- "If you were the hiring manager for this role, what would worry you most about a candidate coming from a smaller company than mine?"
- "What questions would a hiring manager at this company ask that a generic interview guide wouldn't include?"
This is roughly the mirror image of what's happening on the other side of the hiring desk. Recruiters are running their own version of this research using ChatGPT prompts for recruiters to shortlist you before you even get to interview stage, so it only makes sense to do the same digging back.
Prompts to build STAR stories from your actual history
This is the part people skip and it's the part that matters most. STAR stands for Situation, Task, Action, Result, and most people can tell you the acronym but freeze when asked to build one on the spot. Use ChatGPT to mine your own history for stories you'd forgotten you had:
- "Here's my CV [paste it]. For the role of [job title], suggest three past projects from my history that would make strong STAR stories, and tell me which competency each one demonstrates."
- "I'll describe a work situation in messy, unstructured detail. Turn it into a tight STAR story under 150 words. Here's the situation: [paste your rough notes]."
- "Ask me five follow-up questions about this STAR story so I'm ready if the interviewer digs deeper: [paste your STAR story]."
- "Rewrite this STAR story three ways: one focused on leadership, one focused on problem-solving, one focused on working under pressure."
If your CV itself is thin on detail, that's worth fixing before the interview stage even starts. The ChatGPT prompts for resume writing guide covers exactly how to pull stronger, more specific language out of a flat CV, and a sharper CV gives you more raw material to build interview stories from in the first place.
Prompts for the tough, off-script questions
These are the ones that catch people out, not because they're hard, but because they're rarely rehearsed:
- "I'm interviewing for [role] at [company]. Ask me the hardest, most awkward interview question a panel might ask given this job description, and don't soften it."
- "Here's my answer to 'why did you leave your last job.' Tell me honestly if it sounds defensive, and rewrite it so it doesn't."
- "I have a gap in my employment history of 14 months. Help me explain it in one sentence without sounding apologetic."
- "Play the role of a skeptical hiring manager who thinks I'm overqualified for this role. Ask me three pointed questions."
- "What's a question this specific company might ask that's designed to test culture fit rather than skills, based on their careers page?"
The gap-in-employment one deserves a mention on its own. I've had people spend twenty minutes with ChatGPT trying to build an elaborate justification for a nine-month career break, when the honest one-line version, "I took time out to care for a parent, I'm back and focused now," lands far better than a paragraph of spin. Ask the tool to shorten your answer before you ask it to improve it.
Prompts for salary, offer, and negotiation questions
These get asked far earlier in the process than most people expect, sometimes in the first screening call, so having language ready matters:
- "Based on typical salary ranges for [job title] in [city/country], help me draft an answer to 'what are your salary expectations' that gives a range rather than a fixed number."
- "How do I answer 'what's your current salary' when I'd rather not disclose it, without sounding evasive?"
- "Write three versions of a response to a lowball offer, one direct, one diplomatic, one that asks for time to think."
A quick reality check on numbers: as of 2026, UK salary transparency rules mean many job postings now list a band upfront, so use ChatGPT to help you compare that band against Glassdoor and LinkedIn Salary data for the same title and region rather than guessing blind.
The uncomfortable part nobody wants to say out loud
Here's the bit that gets left out of most guides on this topic. If you paste a job description into ChatGPT and ask for the "best possible answer" to a common interview question, you will often get back something that sounds impressive and reads almost identically to what thousands of other candidates are generating from the same prompt this month. Hiring managers who interview regularly have started noticing a pattern, answers that hit every keyword from the job description in a suspiciously tidy order, delivered with the same rhythm and the same three-part structure every time.
I've sat on hiring panels where two candidates in the same week gave what was effectively the same "tell me about a time you handled conflict" answer, right down to the phrase "I scheduled a one-on-one to align on expectations." That's not a coincidence, that's ChatGPT's house style showing through. The fix isn't to avoid the tool, it's to treat its output as a first draft, not a script. Read the answer out loud, then change at least a third of the wording into how you'd say it, with your own filler words, your own pacing, your own slightly messier way of explaining things.
This matters more than it sounds. Interviewers aren't grading you on eloquence, they're listening for whether the story sounds lived rather than composed.
Want AI doing the heavy lifting in your marketing?
I build the systems that handle the boring 80 percent, so you get your week back. Done properly, with the human kept in.
A step-by-step 20-minute prep session
This is the exact sequence I walk clients through the night before an interview:
- Minute 0 to 5: paste the job description into ChatGPT and ask it to list the top five requirements by frequency.
- Minute 5 to 10: paste your CV and ask for three STAR stories matched to those five requirements.
- Minute 10 to 14: ask ChatGPT to play a skeptical interviewer and fire three hard follow-up questions at each story.
- Minute 14 to 17: rewrite the two weakest answers in your own words, out loud, without looking at the screen.
- Minute 17 to 20: ask for one likely salary or offer question specific to the role and draft a one-sentence answer to it.
Twenty minutes, five steps, and you walk in with tailored material instead of a memorised script. If you want to take the practice further and rehearse the actual back-and-forth out loud rather than just reading answers off a screen, the ChatGPT prompts for practising a job interview out loud guide has the exact wording to set that up as a live mock interview.
What to avoid entirely
A few habits I've seen backfire more than once:
- Memorising an answer word for word. It's obvious in the room, and one unexpected follow-up question knocks the whole script over.
- Using ChatGPT to invent achievements or numbers you can't back up if asked to explain them. Interviewers ask "how did you measure that" more often than people expect.
- Asking for "the perfect answer" instead of "an honest answer that fits my actual experience." The first produces polish, the second produces something you can defend under questioning.
- Skipping the out-loud rehearsal. Reading an answer silently and speaking it aloud for the first time in the interview are two very different skills.
If you want more ready-made prompt sets to build on this for other business or job-search situations, there's a running collection of free prompt packs worth checking, covered in where to find a free PDF of ChatGPT prompts for business, along with a note on which ones are worth handing over your email address for.
I keep every related walkthrough in the ChatGPT Prompts and Limits: 30 Guides to Fun Prompts, Photo Edits and Daily Caps.
I built a free ChatGPT prompt generator for exactly this.
Related: fintech: guidelines and how to pitch.
Frequently asked questions
Can interviewers tell if I used ChatGPT to prepare for an interview?
They can't tell you used a tool to prepare, everyone researches beforehand and that's expected, but they can often spot answers that were copied without editing, because the phrasing has a recognisable AI rhythm and hits keywords from the job description a little too neatly. The fix is to rewrite ChatGPT's draft in your own words before you rehearse it.
What's the single best ChatGPT prompt for interview prep?
Paste the full job description plus your CV and ask: "Based on this job description and my CV, list the three requirements I'm weakest on and give me one STAR story for each that addresses the gap honestly." It forces the tool to work from your real material rather than generic advice.
Should I use ChatGPT during a live interview, like on a video call?
No. Beyond the ethics of it, most interviewers can tell when someone is reading or pausing to type, and if you're caught it ends the process immediately. Use ChatGPT for prep only, never as a live crutch during the actual conversation.
How far in advance should I start using ChatGPT to prepare?
Two to three days ahead gives you enough time to research the company, build STAR stories, and rehearse them out loud more than once. Doing it all the night before, as most people do, still works for the twenty-minute session above, but you lose the chance to sleep on your answers and refine them a second time.