Question 20
A reviewer needs a focused diagnosis and proof of resolution for: A developer pasted real customer support transcripts into a Copilot prompt to generate a troubleshooting example. Which check-to-acceptance pair addresses that specific condition?
Explanation
Correct Answer
The read-only check (inspect the submitted prompt and compare its customer fields against the approved data-classification policy) isolates the relevant issue, and the acceptance evidence (the revised prompt contains no customer identifiers and still yields a useful sanitized example) shows that its corresponding fix worked.
Incorrect Answers
This diagnostic and result belong to a different incident: A hiring team considers an AI-generated developer candidate summary that consistently favors candidates from one background. For 'A developer pasted real customer support transcripts into a Copilot prompt to generate a troubleshooting example', examine inspect the submitted prompt and compare its customer fields against the approved data-classification policy and confirm that the revised prompt contains no customer identifiers and still yields a useful sanitized example.
This diagnostic and result belong to a different incident: An AI-drafted error handler catches every exception and suppresses evidence of customer payment failures. For 'A developer pasted real customer support transcripts into a Copilot prompt to generate a troubleshooting example', examine inspect the submitted prompt and compare its customer fields against the approved data-classification policy and confirm that the revised prompt contains no customer identifiers and still yields a useful sanitized example.
This diagnostic and result belong to a different incident: A teammate accepts a Copilot claim that an undocumented flag disables all security checks for a build tool. For 'A developer pasted real customer support transcripts into a Copilot prompt to generate a troubleshooting example', examine inspect the submitted prompt and compare its customer fields against the approved data-classification policy and confirm that the revised prompt contains no customer identifiers and still yields a useful sanitized example.



GitHub Copilot Practice Tests