AI Governance CAIPM Practice Questions (Page 5)

Questions 21–25

1. 21expert · hard

   A multinational corporation operates in multiple jurisdictions with varying AI regulations. The company is developing a global AI governance framework. The program manager must balance the need for consistent standards with local legal requirements. Which approach is most effective?

   • AUse the company's home country regulations as the sole standard for all operations.
   • BCreate separate governance frameworks for each jurisdiction without any global consistency.
   • CDefer all governance decisions to local business units with no central oversight.
   • DAdopt a single global framework based on the most stringent regulations and allow local adaptations where necessary.
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   Which of the following is a widely recognized AI governance framework that emphasizes principles such as transparency, fairness, and accountability?

   • AISO 27001
   • BTOGAF
   • COECD AI Principles
   • DCOBIT
   Reveal Solution Discussion Select an answer first
4. 23application · medium

   A bank has deployed an AI system for fraud detection. The system flags suspicious transactions. The bank's compliance team wants to ensure the system remains effective and compliant. What is the best way to audit the AI system?

   • ARely on the model's performance metrics reported by the data science team
   • BPerform periodic audits that include testing the model on new data and reviewing decision outcomes
   • CReview the model's code and documentation annually
   • DConduct an audit only when a regulatory issue arises
   Reveal Solution Discussion Select an answer first
5. 24expert · hard

   A government agency is developing an AI system for predictive policing. The system has been criticized for potential racial bias. The agency must balance public safety benefits with ethical concerns. The program manager must decide on an approach. Which approach is most balanced?

   • ADeploy the system without bias testing to avoid delays.
   • BUse the system only in low-crime areas to minimize bias.
   • CImplement the system with strict bias testing, continuous monitoring, and community oversight.
   • DCancel the system entirely due to ethical concerns.
   Reveal Solution Discussion Select an answer first
6. 25application · medium

   A multinational bank is deploying an AI system for credit scoring. The system will process personal data of EU citizens. The legal team requires that the bank document how the AI makes decisions and provide meaningful explanations to customers. Which regulatory requirement is directly applicable to this scenario?

   • AThe GDPR's right to explanation and data protection impact assessment requirements.
   • BThe NIST AI RMF, because it is a legally enforceable regulation.
   • CThe HIPAA Privacy Rule, because credit scoring involves financial data.
   • DThe CCPA's requirement to delete all AI training data upon request.
   Reveal Solution Discussion Select an answer first