AI Organizational Processes and Alignment AAIR Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   How do AI-related processes align with an enterprise risk management (ERM) framework?

   • ABy ignoring ERM and focusing only on technical AI risks
   • BBy delegating all AI risk decisions to external auditors
   • CBy replacing the ERM framework with a separate AI-specific risk framework
   • DBy using the same risk categories, assessment methods, and reporting lines as the ERM framework
   Reveal Solution Discussion Select an answer first
2. 17application · medium

   A financial services firm's board audit committee is reviewing the AI risk management process. The committee finds that AI incidents are reported to the IT department but not to the board. The committee wants to ensure it has visibility into material AI risks. Which action best integrates AI risk reporting into corporate governance?

   • AAsk the external auditor to review all AI incidents and report directly to the board.
   • BDefine materiality thresholds for AI incidents and require reporting to the board audit committee when thresholds are met.
   • CHave the AI team present a detailed technical report at each board meeting.
   • DRequire the IT department to send all AI incident reports to the board.
   Reveal Solution Discussion Select an answer first
4. 18foundation · easy

   Which mechanism is commonly used to support continuous improvement of AI risk management processes?

   • ARegular audits and lessons-learned reviews
   • BOne-time risk assessment at project initiation
   • CKeeping processes static to ensure consistency
   • DIgnoring past incidents to avoid blame
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   What is the primary purpose of the oversight process in AI risk management?

   • ATo allocate financial resources for AI initiatives
   • BTo ensure that AI risk management activities are monitored and reviewed for effectiveness
   • CTo design the user interface for AI applications
   • DTo develop and train AI models
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   An insurance firm has deployed an AI underwriting model. Six months after launch, the model's performance degrades due to changing market conditions. The model risk team conducts a review and identifies that the model was not retrained because no feedback loop existed. Which mechanism should the firm implement to prevent recurrence?

   • ARetrain the model immediately and redeploy it without further review.
   • BIncrease the frequency of board reporting on model performance.
   • CReplace the AI model with a traditional statistical model to avoid future issues.
   • DEstablish a periodic model performance review with triggers for retraining based on predefined thresholds.
   Reveal Solution Discussion Select an answer first