ResumeGuard

Answer

AI-generated resume red flags

What are AI-generated resume red flags?

Direct answer

Common AI-generated resume red flags are unfilled placeholders (Jane Doe, [Company Name]), identical bullet length and openers across every job, transformational verbs paired with round percentages and no stack or project names, generic linkedin.com links without a profile slug, and PDF metadata from resume builders. Those signals measure polish — they do not by themselves prove the career is invented. Pair them with employer and timeline checks.

Document and template tells

  • Example identity: Jane Doe, 123 Main St, example@email.com
  • Bracketed tokens left in the PDF: [Your Title], [Company]
  • linkedin.com with no /in/username
  • The same sentence cadence in every role

Language tells

Machine-polished resumes overuse words like spearheaded, orchestrated, and leveraged, plus round metrics (improved efficiency by 40%) without constraints, ticket numbers, or product names. Real resumes usually have uneven phrasing somewhere on page two.

What to do next

Ask the candidate to walk through one named project. Run a structured authenticity check so polish risk and fabrication risk are scored separately. Do not reject solely because a resume looks AI-edited.

Longer read: Longer guide: AI-generated resumes for recruiters

See pricing · Sign up

FAQ

Common questions

Is an AI-written resume automatically fraudulent?

No. Many real candidates use AI to edit. Fraud is invented employers, dates, or credentials. ResumeGuard separates polish risk from fabrication risk for that reason.

Can software detect ChatGPT resumes?

Software can flag writing-pattern and template signals. It cannot prove which model was used. Treat those flags as a reason to ask better interview questions.

How do I screen for this in an ATS?

Call a resume screening API at upload and store polish_risk plus writing-pattern flags on the candidate. See the ATS API page and the public OpenAPI spec.

Still have questions? Contact hello@resumeguard.io