引導式測驗 · 中等 等級

檢測人工智慧撰寫的履歷和候選人詐欺測驗

測試基於證據的驗證、人工智慧寫作偵測器的限制、一致性檢查、身分審查、佐證、隱私和申請人追索權。

問題 1 的 8

A resume is polished and uses generic wording. What does that prove?

預覽每個問題8 問題

當您參加測驗時,就會出現答案和解釋。

  1. A resume is polished and uses generic wording. What does that prove?
    • That an AI detector is correct
    • Only that the text has those characteristics, not that the applicant committed fraud
    • That the candidate lacks the skill
    • That the identity is synthetic
  2. How should an AI-writing detector score be used?
    • As an automatic rejection threshold
    • As a credential verification record
    • As a signal for review, not proof of fraud
    • As a substitute for interviewing
  3. Which check best verifies a required professional license?
    • Ask a model whether the resume sounds genuine
    • Compare the applicant’s punctuation
    • Reject the application if the document is a PDF
    • Confirm it with an appropriate issuing source
  4. Why use the same job-related work-sample follow-up for comparable candidates?
    • Consistent criteria make evidence more comparable
    • It guarantees there is no bias
    • It makes identity checks unnecessary
    • It tests whether candidates used identical wording
  5. A remote interview image contains an unusual artifact. What is a sound response?
    • Accuse the candidate immediately
    • Preserve the evidence and use a documented human review before drawing a conclusion
    • Treat the artifact as conclusive proof of a proxy
    • Publish the screenshot to request opinions
  6. What should a verification policy explain?
    • Only the detector’s name
    • How to evade the detector
    • What is checked, why, and how an applicant can correct an error
    • The recruiter’s private opinion
  7. Why limit identity data collected during hiring?
    • More data always improves accuracy
    • Biometric data cannot be sensitive
    • Collection removes false positives
    • Sensitive verification data should be limited to the legitimate hiring purpose
  8. What should precede adverse action based on a fraud concern?
    • Corroborating evidence and a documented human decision
    • An unreviewed detector alert
    • A recruiter’s intuition alone
    • A change in writing style