AI Strategy and Adoption Roadmap Design CAIPM Practice Questions (Page 8)

Questions 36–40

1. 36application · medium

   A regional bank wants to deploy AI for fraud detection. The initial assessment reveals that transaction data is siloed across three legacy systems, the data science team has only basic Python skills, and the compliance department has not been consulted. The program manager must design the first phase of the adoption roadmap. Which approach best aligns with the readiness assessment findings?

   • ABegin with a small pilot using a single data source, paired with targeted training for the data science team and early compliance engagement.
   • BLaunch a full-scale deployment across all three systems immediately to maximize business value from day one.
   • COutsource the entire AI solution to a vendor to bypass internal skill and data integration challenges.
   • DDefer the AI initiative until the data science team completes a six-month advanced training program.
   Reveal Solution Discussion Select an answer first
2. 37expert · hard

   A financial-services firm is building an AI system that recommends investment portfolios. The model uses customer demographic data. The legal team warns that the model could produce biased outcomes for certain groups. The data science team wants to release the model quickly to meet a quarterly revenue target. The compliance officer requires a documented risk assessment before any customer-facing release. What is the best course of action?

   • ARelease the model on schedule and conduct a retrospective bias audit after three months
   • BHave the legal team approve the model without a technical review
   • CRemove demographic data from the model to eliminate bias concerns and release immediately
   • DDelay the release until a formal bias and fairness assessment is completed and documented
   Reveal Solution Discussion Select an answer first
4. 38expert · hard

   A healthcare company is planning an AI diagnostic-support system. The company has a $1M budget. The data team estimates that data cleaning and integration will cost $400,000, model development $300,000, and deployment $200,000. The compliance team requires a $100,000 independent audit. The CFO wants to reduce the budget by 10%. Which cost item is the most appropriate to reduce?

   • AReduce deployment to $150,000 and delay full integration to a later phase
   • BReduce model development to $200,000
   • CReduce the compliance audit to $50,000
   • DReduce data cleaning to $300,000 to save $100,000
   Reveal Solution Discussion Select an answer first
5. 39application · medium

   A retail company has deployed an AI recommendation engine in its mobile app. The product team wants to know if the system is improving customer engagement. They have access to click-through rate, average order value, and customer churn rate. The AI team wants to know if the model itself is performing well. Which metric is most appropriate for measuring the business impact of the AI system?

   • AClick-through rate on recommended items
   • BNumber of model retraining cycles per month
   • CModel accuracy on a held-out test set
   • DTime spent by data scientists on model maintenance
   Reveal Solution Discussion Select an answer first
6. 40expert · hard

   A global manufacturing firm is designing a three-year AI roadmap. The company has a mature data platform but limited AI talent. The leadership team is divided: the COO wants to prioritize operational efficiency, the CMO wants to prioritize customer personalization, and the CFO insists on a strict budget with no overruns. The program manager must sequence the roadmap. What is the most defensible approach?

   • ADelay all AI initiatives until the company hires a full team of data scientists.
   • BSequence the roadmap to start with operational efficiency use cases that leverage existing data, then expand to customer personalization in later phases.
   • CPrioritize customer personalization first because it has higher revenue potential, even though it requires new data collection.
   • DRun both operational efficiency and customer personalization initiatives in parallel from the start to satisfy all leaders.
   Reveal Solution Discussion Select an answer first