How to use AI in your job search without sounding generic

Start with your own words, not a blank prompt

Before you open any AI tool, write two or three sentences about the job in your own voice. Do not worry about grammar or polish. Just answer: what did you actually do, what changed because of it, and what would you do differently now? These raw notes become the source material you feed the AI.

When you prompt the tool, paste your notes in and say something like: 'Rewrite this in clear professional English, keeping the specific details and the first-person voice.' That instruction alone cuts most of the generic output because the AI has real facts to work with instead of filling gaps with filler.

• Write your own bullet first, then ask AI to tighten it
• Include numbers, team sizes, tools, and timelines in your notes
• Tell the AI explicitly to keep any specific detail you mention

Feed the job posting into every prompt

A generic AI output usually comes from a generic prompt. If you ask AI to 'write a cover letter for a marketing manager role,' it has nothing to anchor to. If you paste the full job posting and your relevant experience together, the output mirrors the employer's language and priorities.

Copy the posting text, strip the company boilerplate, and highlight two or three requirements that match your background. Then prompt: 'Write a cover letter opening paragraph that connects my experience with these specific requirements.' The result will reference the actual role, not a hypothetical one.

• Paste the full posting, not just the job title
• Point the AI to the two or three requirements you match most closely
• Ask it to use the employer's exact terminology where it fits naturally

Edit for your voice after every AI draft

Read the AI draft out loud. If you would not say a sentence in a conversation, cut or rewrite it. Common offenders include 'leveraged synergies,' 'spearheaded initiatives,' and any sentence that starts with 'As a dedicated professional.' Replace them with the plain version of what you mean.

A useful test: swap your name for a colleague's name and read the document again. If it still fits them perfectly, it is not specific enough to you. Add one detail that only applies to your situation, such as the industry you came from, the size of the team you managed, or the constraint you were working under.

1. Read the draft out loud and mark anything you would not say
2. Replace filler phrases with the plain version of the fact
3. Add one detail that is unique to your experience
4. Run the name-swap test before you submit

Use AI to tailor, not to invent

AI is good at reorganising and rephrasing what you give it. It is not good at knowing what you actually did. If you let it fill gaps, it will invent plausible-sounding but vague claims that a recruiter or interviewer will see through immediately.

Keep a plain-text document with your real achievements, the tools you have used, and the outcomes you can speak to. Treat this as your source of truth. Every time you use AI to tailor an application, pull from that document. If the AI suggests a claim that is not in your document, delete it.

How Break-In Bad keeps your applications specific

Break-In Bad's assistant Omar reads the full job posting before writing anything, including location, visa requirements, time zone, and seniority level. She then asks you up to 20 questions about your background through a conversation you can tap or speak, and builds your resume and cover letter from the facts you confirm. Every line is checked against what you said, and your original wording is preserved where it fits.

Because the tailoring starts from your confirmed facts rather than from a generic template, the output reflects your actual experience. You review the result before anything is submitted, so you stay in control of what goes out under your name.

Phrases to cut and what to write instead

Certain phrases appear so often in AI-assisted applications that they now signal to recruiters that the document was not carefully edited. Cutting them and replacing them with specific alternatives takes about ten minutes and makes a measurable difference in how the document reads.

• Cut 'results-driven' and write the result: 'reduced onboarding time from six weeks to three'
• Cut 'passionate about' and write the action: 'have built three internal tools in this space'
• Cut 'strong communication skills' and write the context: 'presented quarterly findings to a board of twelve'
• Cut 'team player' and write the collaboration: 'coordinated across four time zones with engineers in Sydney and Toronto'
• Cut 'detail-oriented' and write the evidence: 'caught a billing error that saved the client $14,000'