AI Pilot Execution and Scaled Deployment CAIPM Practice Questions (Page 2)

Questions 6–10

1. 6expert · hard

   A pharmaceutical company is scaling an AI model that predicts drug interactions. The model was trained on global patient data, but the scaled deployment must comply with GDPR and other regional regulations. The compliance team requires an audit trail of all model decisions. Which approach best satisfies both regulatory and operational needs?

   • AStore all raw patient data in a central repository for auditing purposes.
   • BUse a black-box model with no logging to protect patient privacy.
   • CDeploy the model globally and log all predictions with timestamps and input data hashes.
   • DDeploy the model only in regions with no strict data regulations to avoid compliance overhead.
   Reveal Solution Discussion Select an answer first
2. 7foundation · easy

   What is the primary goal of a scaled deployment strategy?

   • ATo keep the AI solution in pilot mode indefinitely
   • BTo reduce the number of users of the AI solution
   • CTo expand the AI solution from a limited pilot to full production operation
   • DTo delete all pilot data
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   When evaluating pilot outcomes, what is the primary basis for determining if the pilot was successful?

   • AThe subjective opinion of the project sponsor
   • BThe amount of data processed by the AI model
   • CComparison of actual results against the predefined success criteria
   • DThe number of meetings held during the pilot
   Reveal Solution Discussion Select an answer first
5. 9expert · hard

   A telecom company's AI network-anomaly-detection pilot succeeded in one region. The scaled deployment will cover all regions, but each region has different data-privacy laws and network infrastructure. The program manager must design a scalable deployment strategy. Which approach is most effective?

   • ADeploy regional instances of the model with local data processing, and use a central governance framework to ensure compliance.
   • BDeploy a single centralized model that processes all regions' data to ensure consistency.
   • CDeploy the model only in regions with the most advanced infrastructure to simplify scaling.
   • DUse a single cloud provider in one country to host all data and models.
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A government agency is scaling an AI document-classification system from a pilot in one department to the entire agency. Employees are worried about job displacement. The deployment lead must address these concerns while ensuring compliance with records-retention policies. What should the lead do?

   • ADeploy the AI system quietly and let employees discover its benefits over time.
   • BReassure employees that no jobs will be affected, even if that is not certain.
   • CCommunicate that the AI will augment employees' work, not replace them, and provide upskilling opportunities.
   • DFocus only on technical training and ignore emotional concerns.
   Reveal Solution Discussion Select an answer first