Governance, Ethics, and Responsible AI in Adoption CAIPM Practice Questions (Page 4)

Questions 16–20

1. 16expert · hard

   A multinational corporation is adopting AI across multiple business units. Each unit has its own data practices and risk tolerance. The AI program manager must create a unified governance framework. What is the best approach?

   • ADevelop a centralized governance framework with mandatory standards, but allow business units to implement additional controls as needed.
   • BAdopt a single, rigid framework that applies identically to all business units.
   • CLet each business unit develop its own governance framework independently.
   • DOutsource governance to an external consultancy to ensure neutrality.
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A city government is deploying an AI system to triage citizen service requests. The mayor's office wants to announce the system publicly, but the privacy advocacy group has raised concerns about surveillance. The program manager needs to develop a stakeholder engagement strategy. What is the most effective approach?

   • AProceed with the deployment and issue a press release after launch to avoid delaying services.
   • BHold a series of public town halls and create an online feedback portal before finalizing the deployment plan.
   • CLimit communication to the privacy advocacy group and the mayor's office to avoid public confusion.
   • DPublish a detailed technical white paper on the AI model's architecture and training data.
   Reveal Solution Discussion Select an answer first
4. 18expert · hard

   An AI system that recommends job candidates has been found to produce biased outcomes against a protected group. The company's legal team is concerned about regulatory penalties, while the engineering team wants to quickly fix the model. The AI program manager must decide on a course of action. What is the most responsible approach?

   • AImmediately re-train the model with a balanced dataset and redeploy it as soon as possible.
   • BKeep the system running but add a warning to users that results may be biased.
   • CLimit the system's use to demographic groups that are not affected by the bias.
   • DTemporarily halt the system, conduct a thorough bias investigation, and implement corrective measures with ongoing monitoring.
   Reveal Solution Discussion Select an answer first
5. 19application · medium

   An AI system used for hiring has been found to have a disparate impact on a protected group. The company must decide how to address this. What is the most appropriate risk mitigation action?

   • ADiscontinue the AI system and revert to manual hiring processes.
   • BIgnore the disparate impact because the system is more efficient than manual hiring.
   • CConduct a disparate impact analysis and adjust the model's decision threshold to reduce the adverse impact.
   • DKeep the system but provide diversity training to hiring managers.
   Reveal Solution Discussion Select an answer first
6. 20foundation · easy

   Which of the following best describes the purpose of an AI governance framework?

   • ATo guarantee that AI models always produce accurate results
   • BTo automatically enforce all legal regulations without manual intervention
   • CTo replace the need for human decision-making in AI operations
   • DTo provide a structured set of policies, processes, and responsibilities for overseeing AI initiatives
   Reveal Solution Discussion Select an answer first