Sustaining AI Transformation and Continuous Improvement CAIPM Practice Questions (Page 9)

Questions 41–45

1. 41foundation · easy

   Why should an organization consider both quantitative and qualitative benefits when measuring the ROI of an AI initiative?

   • ABecause some benefits, such as improved customer satisfaction, are not easily quantified but still contribute to long-term value
   • BBecause qualitative benefits are always more important than quantitative ones
   • CBecause quantitative benefits are unreliable
   • DBecause qualitative benefits are easier to measure
   Reveal Solution Discussion Select an answer first
2. 42application · medium

   A healthcare AI startup has developed a diagnostic support tool that will be deployed in hospitals. The legal team warns that the tool must comply with medical device regulations and maintain patient privacy. The startup wants to ensure ongoing compliance as the model is updated. What should the startup implement as part of its governance framework?

   • AA one-time regulatory approval before launch, with no further oversight.
   • BA policy that prohibits any model updates after initial deployment to avoid regulatory risk.
   • CA technical solution that encrypts all patient data, which alone ensures compliance.
   • DA documented change management process that includes regulatory impact assessment for every model update.
   Reveal Solution Discussion Select an answer first
4. 43application · medium

   A telecommunications company has a churn-prediction model that is retrained monthly. The model's performance has been stable, but the data science team wants to ensure that the model remains valid as customer behavior evolves. They also need to maintain a clear history of changes for internal audits. What should the team implement?

   • AA scheduled retraining pipeline that automatically logs each model version, its training data, and evaluation metrics.
   • BA policy to retrain only when the model's accuracy drops below 80%.
   • CA manual retraining process where the data scientist saves the model file with a new name each month.
   • DA dashboard that shows the current model's performance but does not store historical versions.
   Reveal Solution Discussion Select an answer first
5. 44application · medium

   A media company uses an AI content-recommendation system. The system has increased user engagement by 10%, but the company is unsure whether the increase is sustainable. The data science team wants to add new features to the model, but the product team is concerned about user privacy. What should the program manager do to balance these concerns?

   • AProceed with the new features because engagement is the primary goal
   • BAbandon the new features to avoid any privacy risk
   • CConduct a privacy impact assessment before implementing the new features
   • DImplement the features and monitor engagement only
   Reveal Solution Discussion Select an answer first
6. 45expert · hard

   A large insurance company has multiple AI initiatives across different departments. Each department manages its own AI projects with its own vendors and tools. The company wants to sustain these initiatives and ensure they align with corporate strategy, but the departments resist central oversight. What is the best governance approach to balance autonomy and alignment?

   • ACentralize all AI decision-making under a single AI office to ensure consistency.
   • BAllow each department to continue independently and share results in an annual report.
   • CMandate that all departments use the same vendor and toolset to simplify governance.
   • DEstablish a federated governance model with a central AI council that sets standards and local units that manage execution.
   Reveal Solution Discussion Select an answer first