nerdexam
Microsoft

GH-300 · Question #29

As a developer who is evaluating the efficiency of GitHub Copilot in generating accurate code across different problem domains. During your R&D, you want to consider the limitations of GenAI tools…

The correct answer is C. Copilot may return output that reflects bias or inaccurate info because of its training dataright. Copilot is trained on a vast amount of data, which is publicly available; this data may include the biased, outdated, or incorrect practices of code implementation. This typically highlights a key concern in Responsible AI…

Responsible AI

Question

As a developer who is evaluating the efficiency of GitHub Copilot in generating accurate code across different problem domains. During your R&D, you want to consider the limitations of GenAI tools like Copilot. Which of the following options best describes a core limitation?

Options

  • ACopilot always generates error-free code across all languages
  • BCopilot reduces the need for STLC and code reviews
  • CCopilot may return output that reflects bias or inaccurate info because of its training dataright
  • DCopilot can only be used offline without internet access

How the community answered

(23 responses)
  • B
    9% (2)
  • C
    87% (20)
  • D
    4% (1)

Explanation

Copilot is trained on a vast amount of data, which is publicly available; this data may include the biased, outdated, or incorrect practices of code implementation. This typically highlights a key concern in Responsible AI. https://learn.microsoft.com/en-us/training/modules/responsible-ai-with-github-Copilot/2-manage-

Topics

#Copilot limitations#AI bias#Inaccurate output#Responsible AI use

Community Discussion

No community discussion yet for this question.

Full GH-300 Practice