Emerging AI Trends and Future Opportunities CAIPM Practice Questions (Page 10)

Questions 46–49

1. 46expert · hard

   A global bank wants to use generative AI for personalized financial advice. The bank operates in multiple countries with varying data residency laws. The compliance team requires that all customer data remain in the country of origin. The bank has a strong AI team but limited experience with cross-border data governance. What is the most strategically sound approach?

   • AAbandon the project because generative AI cannot comply with data residency laws.
   • BDeploy separate generative AI models in each country, using local data for training and inference.
   • CDeploy a single global model that processes all data in a central location.
   • DUse a third-party cloud provider that stores all data in a single offshore location.
   Reveal Solution Discussion Select an answer first
2. 47expert · hard

   A healthcare provider wants to deploy a generative AI system to draft patient discharge summaries. The system will be trained on historical records that contain demographic biases. The provider operates in a jurisdiction with strict data residency laws and has a risk-averse legal team. The clinical staff are concerned about liability. What is the most critical risk to address before piloting this system?

   • AThe model may produce biased summaries that lead to unequal care recommendations for different demographic groups.
   • BThe system will require continuous internet connectivity, which is not available in all hospital wards.
   • CThe legal team will reject the project because generative AI is universally prohibited in healthcare.
   • DThe data residency law requires all processing to occur on-premises, which the current cloud provider cannot support.
   Reveal Solution Discussion Select an answer first
4. 48expert · hard

   A logistics company wants to create a roadmap for adopting autonomous agents and edge AI. The company has a moderate budget and a risk-averse culture. The company also faces regulatory scrutiny regarding data privacy and safety. What is the most important element to include in the roadmap to manage uncertainty?

   • AA detailed plan for a single large-scale deployment at the end of the five-year period.
   • BA phased plan with clear go/no-go decision points that incorporate regulatory compliance reviews and pilot results.
   • CA plan that focuses on edge AI only, avoiding autonomous agents due to their complexity.
   • DA plan that commits to a specific technology vendor for the entire five-year period.
   Reveal Solution Discussion Select an answer first
5. 49application · medium

   A healthcare provider wants to use generative AI to draft patient discharge summaries. The provider has a strong IT infrastructure and a culture that encourages innovation, but the clinical staff are concerned about accuracy and liability. The AI program manager must address these concerns before deployment. What is the most appropriate action?

   • ACancel the project because generative AI is too risky for healthcare.
   • BDeploy the system with automatic finalization to save clinician time and increase efficiency.
   • CImplement a human-in-the-loop review process where clinicians verify all AI-generated summaries before they are finalized.
   • DLimit the AI to generating summaries only for low-risk patients to reduce liability.
   Reveal Solution Discussion Select an answer first