Governance, Ethics, and Responsible AI in Adoption CAIPM Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   An AI program manager is leading the deployment of a predictive maintenance system for a manufacturing plant. The system uses sensor data to predict equipment failures. The plant manager wants to minimize downtime, but the data science team warns that the model has a high false-positive rate. The program manager must balance operational efficiency with model reliability. What is the best approach?

   • ADeploy the model without any review process to maximize automation and speed.
   • BImplement a human-in-the-loop review process where maintenance crews verify alerts before acting.
   • CDisable the model until the false-positive rate is zero.
   • DIncrease the model's sensitivity to catch more failures, accepting more false positives.
   Reveal Solution Discussion Select an answer first
2. 2application · medium

   A financial services firm is adopting an AI system for fraud detection. The firm operates in multiple countries with varying regulations. The AI program manager needs to establish a governance framework. Which component is most critical to include?

   • AA detailed technical specification of the AI model's architecture.
   • BA cross-functional governance committee with representation from legal, compliance, data science, and business units.
   • CA single compliance officer responsible for all AI-related decisions.
   • DA policy that all AI systems must be approved by the CEO before deployment.
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   What is the primary goal of bias mitigation in responsible AI practices?

   • ATo ensure AI models are trained on the largest possible datasets
   • BTo make all AI decisions fully automated without human review
   • CTo eliminate all differences in model predictions
   • DTo reduce or eliminate unfair bias in AI systems
   Reveal Solution Discussion Select an answer first
5. 4application · medium

   An e-commerce company is building a recommendation system. The program manager wants to embed governance across the AI lifecycle. The team is currently in the data preparation phase. Which action best integrates governance at this stage?

   • ADefer all governance activities until after the model is deployed to avoid slowing development.
   • BAsk the legal team to review the final model's outputs for compliance before launch.
   • CSelect the model algorithm now to ensure it is explainable, ignoring data issues.
   • DDocument data sources, assess data quality, and check for potential biases in the training data.
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A healthcare AI startup is developing a diagnostic tool. The tool has shown high accuracy in internal tests but has not been validated on diverse populations. The startup is under pressure to launch quickly to secure funding. The AI program manager must balance speed with responsible AI. What is the best course of action?

   • ALaunch the tool immediately and monitor for bias after deployment.
   • BLaunch the tool only in regions with similar demographics to the training data.
   • CConduct a limited pilot with diverse patient groups and use the results to refine the model before full deployment.
   • DDelay the launch indefinitely until the model is validated on every possible demographic.
   Reveal Solution Discussion Select an answer first