AI Best Practices XK0-006 Practice Questions (Page 5)

Questions 21–23

1. 21foundation · easy

   A team notices that an AI code generator produces variable names and comments that reflect a single cultural perspective. Which action best addresses this bias?

   • AUse the generated code as-is to avoid introducing human bias.
   • BReview and replace biased terms with neutral, inclusive alternatives.
   • CSwitch to a different AI tool without reviewing its outputs.
   • DIgnore the issue because code functionality is unaffected.
   Reveal Solution Discussion Select an answer first
2. 22application · medium

   An AI assistant generated a Bash script that uses `curl` to download a file from an internal server and then executes it. The script is intended for a production environment. What is the MOST important security concern?

   • AThe script does not include a progress bar for the download.
   • BThe script is not idempotent and may fail if run multiple times.
   • CThe script uses `curl` instead of `wget`.
   • DThe script downloads and executes code without verifying its integrity or source.
   Reveal Solution Discussion Select an answer first
3. 23application · medium

   A team uses an AI tool to generate regular expressions for log parsing. The tool consistently produces patterns that fail on log entries containing non-English characters, even though the logs are known to be multilingual. The team wants to reduce this bias in the AI's output. What is the BEST approach?

   • APost-process the AI output to replace all non-ASCII characters with '?' before using the regex.
   • BSwitch to a different AI tool that is known for better performance on English text.
   • CInclude examples of multilingual log lines in the prompt and explicitly request patterns that handle Unicode characters.
   • DManually review and fix the regex after every generation, without changing the prompt.
   Reveal Solution Discussion Select an answer first