Using AI to Write Your Resume Without It Sounding Fake
Split the work correctly between you and AI, prompt with facts first, then strip out AI-sounding phrases. Includes a sample prompt and cautions.
You type into ChatGPT: write me a resume for a marketing role. The result looks polished, full of passionate, dynamic and results-driven. But read it again and not one sentence says what you actually did. Recruiters who read dozens of resumes a day see these phrases so often they recognise them instantly.
Why AI resumes sound fake
The problem isn't AI, it's how it's used. When AI writes from scratch, it has none of your real information, so it fills the gaps with generic words that fit anyone. The result sounds good but has no evidence.
Signals HR tends to notice
- Lots of adjectives, no tangible results
- Every bullet starting with the same sentence pattern
- Skills that don't match the experience described
- Language in the interview very different from the resume
Split the work correctly
Your job: provide the facts, including the work you did, tools you used, problems you solved, results and any numbers you genuinely know.
AI's job: organise, tighten wording, find sharper verbs, compare with the job ad for missing keywords, and translate Thai to English.
If you swap roles and let AI invent your achievements, that's where the resume turns fake.
Steps for using AI without drifting from the truth
1. Write the facts in plain language first
No need to be polished, e.g. ran the page for a three-branch coffee shop, made all the posts myself, tried review clips for new menu items, more people messaged asking about the menu then, used Canva and Meta Business Suite.
2. Use a prompt that forces it to stick to facts
Here is my work information [paste] and here is the job ad [paste JD]. Write resume bullets, no more than 4 per role. Use only the information provided. Do not add numbers or achievements that are not in it. If a bullet should have a number but doesn't, put [ ] for me to fill in.
The last line is crucial, because it stops AI from making up numbers.
3. Check and rewrite in your own voice
- Delete empty adjectives like passionate, dynamic or innovative unless backed by evidence.
- Read every bullet aloud. If you couldn't explain it in an interview, delete or fix it.
- Vary sentence structures so every bullet doesn't start the same way.
4. Use AI as a reviewer, not a writer
When you're done, ask AI to play recruiter, e.g. read this resume and tell me which points are vague and what you'd want to ask. The answers also help you prepare for the interview.
Cautions
- Personal data: don't paste your ID number, full address or confidential data from past employers, such as internal sales or client names, into an AI chat.
- Factual errors: AI may get tool names or job titles wrong. Check every proper noun.
- The employer's policy: some companies state conditions on AI use in applications. Read the ad carefully.
- Consistency with LinkedIn: titles, dates and achievements should match everywhere.
Let AI help with the language, but the achievements must be yours.
Before and after
AI from scratch: Results-driven marketer with a passion for creating engaging content that drives growth
AI organising your facts: Managed social media for a 3-branch coffee chain, producing all posts in Canva and testing menu review videos that increased customer inquiries during the campaign
The second isn't fancier, but it's more believable, and you can talk about it right away in an interview.
Start today
- Write the facts of each job in plain language, at least five lines per job.
- Use a prompt that forces real data, and attach the job ad.
- Remove empty words and fill gaps with real numbers.
- Have AI play recruiter for another review.
- Read the whole thing aloud before sending.
Read more on why recruiters are starting to distrust AI-written resumes and adding numbers and results to your resume. Then check your new version with Resume Lab, or ask the AI Career Assistant about tailoring your resume to your field.