Using ChatGPT to Write a Resume That Gets Interviews (2026)

Most CVs written with ChatGPT are worse than the originals. Not because ChatGPT cannot write - it writes fluently - but because people ask the wrong question. They paste their existing CV and say "make this better". What comes back is a polished version of the same generic text, padded with corporate phrases and parallel structures that scream "AI". Recruiters smell it instantly.

This guide shows you how to use ChatGPT properly: as a partner that helps you extract real achievements, write them sharply, tailor them to a specific role, and pass through ATS systems without tripping the obvious AI flags. With the right workflow, you should hit a roughly two-fold improvement in interview rate within two weeks. With the wrong workflow, your CV ends up in the same generic pile as everyone else's.

Why Most CVs Written with ChatGPT Fail

Three reasons, in descending order of severity.

1. Lazy input. If you give ChatGPT a vague brief like "rewrite my CV", it has nothing to work with. It hallucinates plausible-sounding achievements, leans on industry-standard phrases, and produces output that could belong to anyone in your field.

2. No targeting. Generic CVs lose to tailored CVs. ChatGPT can tailor brilliantly to a specific job description - but only if you give it that job description. Most people skip this step.

3. Telltale phrasing. Recruiters in 2026 read 30+ CVs a day. They have learned the AI tells: "leveraged synergies", "spearheaded cross-functional initiatives", "results-driven professional", suspiciously balanced bullet points where every line has the same length. None of these are wrong on their own; together they pattern-match to "wrote this in ChatGPT and did not edit".

The 4-Step CV-with-AI Workflow

Each step has one job. Do not collapse them. The point is to keep your hand on the wheel; ChatGPT is the sat-nav, not the driver.

1. Extract your wins. ChatGPT helps you remember and articulate what you actually achieved.
2. Rewrite for the role. ChatGPT tailors that material to the specific job description.
3. ATS-proof the formatting. ChatGPT checks structure, keywords, and parsing.
4. Generate the cover letter. One last pass tailored to the company.

Total time: 60-90 minutes for a fully tailored application. Faster than doing it yourself, and the output is sharper because you start with raw truth and refine, rather than starting with bland copy and trying to dress it up.

Prompt 1: Extract Your Wins

Open ChatGPT and use this prompt verbatim. Replace the role and timeframe with your own.

You are an experienced career coach interviewing me to surface concrete achievements I can use on my CV. I have been a [JOB TITLE] at [COMPANY] for [TIMEFRAME]. Ask me one question at a time. After my answer, ask a sharper follow-up if my answer was vague (no numbers, no comparison, no stakes). Do not move on until I have given you one specific, measurable accomplishment per major project or responsibility. We are aiming for 8-12 wins. Start now.

Then answer truthfully. The trick is the follow-up loop: when you say "I improved customer satisfaction", ChatGPT pushes back with "by what measure, from what baseline, over what period?" That is the questioning a good career coach would do, and it forces you to articulate things you already did but had not put into language.

By the end of this conversation you should have a list of 8-12 achievements in plain English with numbers. Save it. This is the raw material for everything that follows.

Prompt 2: Rewrite for the Role

Now find the job description for the role you want and use this prompt.

I am applying for the role below. Below the job description is my list of raw achievements. Your job is to rewrite each achievement as a CV bullet that:

1. Starts with a strong action verb (varied - never repeat the same verb twice in a section).
2. Includes a measurable outcome.
3. Maps to a specific requirement or competency in the job description.
4. Reads in plain English, not corporate jargon.

Banned words and phrases: leveraged, spearheaded, results-driven, cross-functional, best-in-class, synergies, value-add, robust, dynamic, passionate.

Output: 6-8 bullets per role, ranked by relevance to the job description. After each bullet, in italics, name which JD requirement it maps to.

JOB DESCRIPTION:
[paste full JD]

RAW ACHIEVEMENTS:
[paste your list from prompt 1]

The "banned words" line alone is worth the whole post. It removes the dominant AI tells in one stroke. The mapping line is what makes the CV demonstrably tailored - when an experienced recruiter reads a bullet that addresses a specific JD requirement, they pause. That pause is what gets you to interview.

A before-and-after example

Before (raw input): "I helped reduce customer churn during a difficult product transition."

After (tailored bullet): "Cut churn from 8.2% to 5.4% during the migration to v3, by redesigning the onboarding email sequence and running 11 customer interviews to surface friction points." Maps to JD requirement: 'experience improving customer retention through data-led iteration'.

That is what good output looks like. Specific, mapped, and human.

Prompt 3: ATS-Proof Formatting

Most ATS systems in 2026 are smarter than they were five years ago, but the basic rules still hold. The following prompt does the structural pass.

Below is my CV draft. Audit it for ATS compatibility and produce a final clean version with these rules:

1. Single column, plain text where possible. No tables, text boxes, headers, footers, or graphics.
2. Standard section headings: Profile, Experience, Education, Skills, Certifications.
3. Standard fonts only (Calibri, Arial, Helvetica).
4. Dates in MMM YYYY format.
5. Job titles, company names, and dates in the same order on every entry.
6. Skills listed as a clean comma-separated list, not a graphic.
7. No emojis, no unicode bullet shapes that some ATS systems break on (use standard hyphens or asterisks).
8. Filename suggestion: FirstName-LastName-CV-Role.pdf

Also: extract the top 12 keywords from the JD I provided in the last message and tell me which ones are missing from the CV. Suggest natural ways to include them.

CV DRAFT:
[paste]

The keyword check is the highest-leverage part. ATS systems still match on literal keywords. If the JD says "stakeholder management" and your CV says "looked after senior partners", you will be filtered out by an ATS that does not understand that those are the same skill. Use the exact phrasing from the JD where it is honest.

Prompt 4: Tailored Cover Letter

Most cover letters are weak because they are abstract. The fix is forcing specifics.

Write a cover letter for the job description above. Constraints:

1. Maximum 280 words.
2. First paragraph: one sentence about why this specific company (not the role - the company). Reference something concrete: a product launch, a strategic shift, a public statement, an industry position. Use the company website I am pasting below.
3. Second paragraph: two specific achievements from my CV that map directly to the top two JD requirements, with numbers.
4. Third paragraph: one sentence about what I want to learn or contribute over the next year.
5. Sign-off: warm but not gushing.
6. Banned phrases: "I am writing to apply for", "I would be a great fit", "passionate about", "I believe my skills".

COMPANY ABOUT-PAGE TEXT:
[paste the About / Mission section]

The "specific company" line is what makes the cover letter read as if you wrote it for them rather than for the entire LinkedIn jobs feed.

Common Mistakes

• Letting ChatGPT invent achievements. If you do not feed it real wins, it will fabricate plausible ones. This will get caught at interview and ruin you. The wins must be real; ChatGPT only helps you articulate them.
• Skipping the JD. A CV without targeting is a generic CV. Use the actual job description every time.
• Submitting the first draft. Always edit. Read it aloud. If a sentence feels stiff, rewrite it in your own register. Your goal is a CV that sounds like you on a confident day, not like a chatbot.
• Going long. Two pages is the UK standard. ChatGPT will pad if you let it. Constrain the prompt: "Maximum 6 bullets per role, maximum 2 pages total."
• Forgetting to update LinkedIn. If your CV says one thing and your LinkedIn says another, recruiters notice. Run the new bullets through LinkedIn too.

UK-Specific Notes

A few details that matter if you are applying in the UK.

• Spelling. "Organisation", "specialise", "behaviour", "centre", "favour". UK English. ChatGPT will default to US English unless you tell it. Add "Use UK English spelling throughout" to every prompt.
• Photo. Standard practice in the UK is no photo on the CV. Some sectors differ; if in doubt, leave it off.
• Date of birth. Do not include it. Standard for both UK and EU.
• Right to work. If you are a non-citizen, a one-line statement near the top ("Right to work in the UK - Skilled Worker visa, expires Aug 2028") saves recruiters a follow-up email.
• References. Do not list references. "References available on request" is also unnecessary in 2026 - assumed.
• Salary expectations. Do not put on CV. Discuss in conversation.

What to Do Next

Run the four prompts above for one application this week. Time it. The first pass will feel slow because you are learning the workflow. By the third application it will take you under an hour.

For more on prompt structure and how to keep your AI outputs sharp, see our guide to 50 best AI prompts for beginners and our piece on how to use AI at work.

Frequently Asked Questions

Can recruiters detect a ChatGPT-written CV?

They detect raw, unedited output - the giveaway phrases, the suspiciously perfect parallel structure. They cannot detect a CV where you used ChatGPT to extract wins, draft phrasing, and tighten language, then edited it into your own register.

Should I use ChatGPT, Claude, or Gemini?

Claude tends to write more naturally for narrative bullets. ChatGPT is faster at structured rewrites. Gemini is fine if you live in Google Docs. For CVs, Claude or ChatGPT - both is better.

How do I avoid generic outputs?

Feed it specifics. Numbers, project names, what was at stake, what you actually did. Banning a list of giveaway phrases in the prompt also helps.

Is it ethical to use AI for my CV?

Yes - provided the achievements are real and the voice is genuinely yours. Using AI to write better-phrased bullets about real work is no different to hiring a CV writer or asking a friend to edit it.

Written by Hus Sultan. For the broader workplace context, see how to use AI at work and our piece on how AI is changing the job market in 2026.