QuestionQ11

Domain III — Understanding how to govern AI development

Which consideration is most important for mitigating potential bias in training and testing data?

  • A Using privacy-enhancing tools to protect the data.
  • B Validating that the data is representative.
  • C Receiving explicit consent to use the data.
  • D Assessing the sufficiency of the third-party data.
Explanation

Representative training and testing data helps prevent systematic underrepresentation or overrepresentation of relevant groups, conditions, and outcomes, reducing biased model behavior.

Community Discussion

No comments yet. Be the first to start the discussion!