Understand Responsible AI Principles GH-300 Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A company is developing a GitHub Copilot extension that generates code for a recruitment platform. The extension is trained on historical hiring data, which contains biases against certain demographic groups. The company wants to ensure fairness while maintaining the efficiency of the tool. What is the most effective approach?

   • AUse a diverse and representative dataset and regularly audit the model for bias.
   • BRely on Copilot's built-in fairness filters.
   • CRemove all demographic data from the training set.
   • DImplement a human review process for all generated code.
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A developer is using GitHub Copilot to generate SQL queries for a database that contains sensitive customer data. Copilot suggests a query that selects more columns than necessary. What is the primary risk in this scenario?

   • ALack of domain expertise
   • BBias
   • CData leakage
   • DHallucination
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   Which ethical principle requires that organizations take responsibility for the outcomes of their AI systems?

   • ATransparency
   • BPrivacy
   • CFairness
   • DAccountability
   Reveal Solution Discussion Select an answer first
5. 4application · medium

   A team is using GitHub Copilot to generate documentation for a new API. They are concerned that the generated documentation might contain inaccurate examples. What is the most effective mitigation strategy?

   • AHave a technical writer review and test all examples.
   • BUse a diverse dataset to train the model.
   • CRely on Copilot's built-in accuracy filters.
   • DUse the documentation as-is because Copilot is trained on official docs.
   Reveal Solution Discussion Select an answer first
6. 5foundation · easy

   A developer asks a generative AI tool for the latest security best practices. The tool provides a response. What is a key limitation to consider?

   • AThe response may be based on outdated information due to the model's training cutoff.
   • BThe response is guaranteed to be free of bias.
   • CThe response will include citations to authoritative sources.
   • DThe response is always up-to-date because the model is continuously trained.
   Reveal Solution Discussion Select an answer first