Governance, Ethics, and Responsible AI in Adoption CAIPM Practice Questions (Page 10)

Questions 46–49

1. 46expert · hard

   An autonomous vehicle company is deploying an AI system for pedestrian detection. The system has a known trade-off between false positives (unnecessary braking) and false negatives (missing pedestrians). The safety team demands zero false negatives, while the engineering team argues that zero false negatives will make the vehicle undrivable due to constant braking. The program manager must reconcile these conflicting requirements. What is the best approach?

   • ASet the detection threshold to minimize false negatives, accepting that the vehicle will brake frequently.
   • BUse a more complex model that eliminates both false positives and false negatives.
   • CLet the engineering team decide the threshold because they understand the system best.
   • DSet the threshold to balance false positives and false negatives based on a risk assessment that considers both safety and drivability.
   Reveal Solution Discussion Select an answer first
2. 47foundation · easy

   Which practice best supports embedding governance into the AI development lifecycle?

   • APerforming a single audit after deployment
   • BRelying solely on automated compliance tools
   • CDocumenting governance only when required by law
   • DConducting ethical reviews at each major milestone
   Reveal Solution Discussion Select an answer first
4. 48foundation · easy

   Which responsible AI practice involves keeping a human in the loop for critical decisions?

   • AHuman oversight
   • BPerformance monitoring
   • CModel versioning
   • DData anonymization
   Reveal Solution Discussion Select an answer first
5. 49application · medium

   A retail company is implementing an AI system for personalized marketing. Privacy advocates have raised concerns about data usage. The AI program manager needs to communicate the company's ethical approach. What is the most effective communication strategy?

   • AIssue a press release highlighting the benefits of personalization without mentioning data usage.
   • BLimit communication to legal disclaimers required by law.
   • CIgnore the concerns and proceed with the system as planned.
   • DPublish a transparent privacy policy and engage with privacy advocates in a dialogue about the system's safeguards.
   Reveal Solution Discussion Select an answer first