AI Use Case Identification and Value Prioritization CAIPM Practice Questions (Page 1)

Questions 1–5

1. 1application · medium

   A financial services firm is prioritizing two AI use cases: (1) an automated loan-approval model and (2) an internal document-classification tool. The loan model has higher projected ROI but uses sensitive customer data and could produce biased outcomes for protected groups. The document tool has moderate ROI and low risk. The compliance officer has raised concerns about regulatory scrutiny. How should the program manager weigh these factors in the prioritization decision?

   • ARun both in parallel, because the compliance officer's concerns can be addressed after the models are built.
   • BPrioritize the document tool first, because its lower risk profile allows the team to build AI capability and trust while the loan model undergoes additional ethical review.
   • CCancel the loan model entirely, because any AI that uses sensitive customer data is too risky to pursue.
   • DPrioritize the loan model, because its higher ROI justifies the additional compliance effort and risk.
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Which activity is part of evaluating risk and ethical considerations for an AI use case?

   • AReviewing the use case for potential privacy violations
   • BCalculating the expected payback period
   • CDetermining if the required data is available
   • DChecking if the use case aligns with the company mission
   Reveal Solution Discussion Select an answer first
4. 3application · medium

   A media company is using a scoring model to rank AI use cases. The model includes criteria: ROI (30%), strategic alignment (25%), feasibility (25%), and risk (20%). Use case X scores high on ROI but low on feasibility. Use case Y scores moderate on all criteria. Which use case should rank higher?

   • AUse case Y, because its balanced scores across all criteria result in a higher weighted total
   • BUse case X, because high ROI is more important than feasibility in the long run
   • CBoth, because they are equally valuable
   • DUse case X, because ROI has the highest weight
   Reveal Solution Discussion Select an answer first
5. 4expert · hard

   A telecommunications company has prioritized three AI use cases: (1) network fault prediction, (2) customer churn prediction, and (3) automated network configuration. The fault prediction model requires real-time telemetry data that is currently being collected but not stored. The churn model requires customer data that is available but has privacy restrictions. The configuration automation requires the fault prediction model to be reliable first. The company has a limited budget and must show value within one year. The data engineering team can build only one new data pipeline in the first quarter. What is the most effective roadmap sequencing?

   • ABuild all three data pipelines in parallel, because the data engineering team can be augmented with contractors.
   • BBuild the telemetry pipeline and fault prediction model first, then use its outputs to enable configuration automation, while deferring churn prediction to a later phase.
   • CPrioritize churn prediction first, because customer data is already available and the model can be built quickly without new infrastructure.
   • DStart with configuration automation, because it has the highest operational impact and can be implemented with rule-based logic initially.
   Reveal Solution Discussion Select an answer first
6. 5application · medium

   A regional bank wants to reduce loan-approval time. The operations team suggests using AI to automate document verification, while the risk team proposes AI for fraud detection. The bank has limited labeled historical data for fraud cases but abundant document images and clear business rules for verification. Which use case should the AI program manager prioritize first?

   • AFraud detection, because it addresses a higher-risk area and will generate more long-term value
   • BNeither, because the bank should first build a comprehensive AI strategy before selecting any use case
   • CBoth in parallel, because they are independent and can be implemented without resource contention
   • DDocument verification, because it has higher data feasibility and clearer success metrics
   Reveal Solution Discussion Select an answer first