Where AI genuinely helps
Resume editing for clarity. Paste your bullets, ask for sharper phrasing, accept what improves and reject what changes meaning.
Generating practice interview questions. "Give me 5 behavioral questions a [role] interviewer at [company] would actually ask." Surprisingly good.
Researching companies. Ask the model to summarize a company’s recent news, products, and culture from public sources. Verify before quoting.
Drafting outreach to networks of 100+ people you genuinely have connections to. Personalize each one before sending.
Where AI hurts you
Writing your cover letter from scratch. Recruiters can spot AI-generated cover letters in 2 seconds in 2026. They read as bland and indistinguishable from every other applicant who used the same prompt. Worse than having no cover letter.
Generating mass applications. Volume-spray applying to 200 jobs is less effective in 2026 than applying thoughtfully to 30. The signal-to-noise has gotten worse.
Letting the model invent things on your resume. Don’t. You will be asked about everything on your resume in the interview.
A simple workflow
Write your draft yourself. Yes, this part is hard. Do it anyway.
Use AI as an editor: paste your draft, ask "what’s unclear here?" and "what’s missing?" Iterate based on the answers.
For company-specific applications, give the AI the job description and your draft. Ask "what specific bullets in my resume should I emphasize for this role?" and "what’s a relevant experience I might be downplaying?" Add or rephrase based on the answer — but in your own voice.
Read everything aloud before sending. If it sounds like AI, it is.