Identify AI Adoption Risks CAIPM Practice Questions (Page 6)

Questions 26–30

1. 26expert · hard

   A financial services company is developing an AI system for credit risk assessment. The model has been found to produce biased outcomes against a minority group. The company is subject to strict regulatory oversight and faces potential fines if the bias is not addressed. The project is behind schedule, and the team is under pressure to deploy. Which action best balances the need to meet the deadline with ethical and legal obligations?

   • ADeploy the model but add a disclaimer that it may have biases.
   • BDelay the deployment until the bias is mitigated and the model is validated for fairness.
   • CRemove the minority group's data from the model to eliminate bias.
   • DDeploy the model now and plan to fix the bias in a future update.
   Reveal Solution Discussion Select an answer first
2. 27expert · hard

   A startup is developing an AI-powered customer service chatbot. The startup has limited funding and needs to achieve a positive ROI within 18 months. The program manager has identified that the chatbot's development costs could overrun due to the complexity of natural language processing. Additionally, the chatbot's performance may not meet customer expectations, leading to low adoption. The startup must decide whether to invest in a more advanced AI model or a simpler rule-based system. Which approach best manages the financial risk?

   • AStart with a rule-based system to minimize development costs and validate customer demand before investing in advanced AI.
   • BInvest in the advanced AI model immediately to ensure the best customer experience and maximize adoption.
   • CSeek additional venture capital funding to cover potential cost overruns and ensure the advanced AI model is built.
   • DOutsource the chatbot development to a low-cost vendor to reduce development costs, but do not test the market.
   Reveal Solution Discussion Select an answer first
4. 28application · medium

   A financial services firm is planning to implement an AI system for fraud detection. The program manager has identified several risks: (1) the model may have a high false positive rate, (2) the system integration with existing transaction processing may be complex, (3) there is a risk of regulatory non-compliance if the system fails to detect certain fraudulent patterns, and (4) the project may exceed its budget. The firm has limited resources and needs to prioritize risks for mitigation. Which risk should be prioritized first?

   • ARegulatory non-compliance, because the impact could include fines and loss of operating license, which are severe and likely if the system fails.
   • BBudget overrun, because the firm has limited resources and any overrun could jeopardize other projects.
   • CHigh false positive rate, because it directly affects customer experience and could lead to lost business.
   • DComplex system integration, because it could delay the project and increase costs.
   Reveal Solution Discussion Select an answer first
5. 29foundation · easy

   During the planning phase of an AI initiative, a program manager wants to systematically identify potential risks that could arise from adopting AI technologies. Which activity best supports this goal?

   • ADeferring risk analysis until after the AI system is deployed to avoid slowing down development
   • BFocusing only on technical risks because they are the most important category
   • CConducting a structured risk identification workshop with stakeholders from technical, legal, and business teams
   • DRelying solely on the vendor's marketing materials to identify risks
   Reveal Solution Discussion Select an answer first
6. 30application · medium

   A European financial services firm is adopting an AI system that processes customer data for credit scoring. The system will be hosted on a public cloud platform with servers located outside the EU. The compliance officer is concerned about GDPR data transfer regulations. Which risk category does this concern primarily fall under, and what is the most appropriate mitigation?

   • ATechnical risk; mitigate by improving the model's accuracy.
   • BOperational risk; mitigate by retraining the staff on the new system.
   • CLegal and compliance risk; mitigate by ensuring data residency in an EU region or implementing appropriate safeguards for transfers.
   • DEthical risk; mitigate by making the AI's decision-making more transparent.
   Reveal Solution Discussion Select an answer first