AI Risk Management AAIA Practice Questions (Page 2)

Questions 6–10

1. 6application · medium · select all that apply

   An insurance company uses AI to automate claims processing. The risk assessment has identified a risk of the AI system incorrectly denying valid claims, leading to regulatory fines and customer dissatisfaction. The company wants to implement risk response strategies. Which of the following are appropriate risk response strategies? (Select all that apply.)

   • APurchase an insurance policy to cover potential regulatory fines.
   • BIgnore the risk because the AI system has been approved by internal IT.
   • CDiscontinue the AI system and process all claims manually.
   • DImplement a human review process for all AI-denied claims before final decision.
   • EAccept the risk because the AI system is faster and cheaper than manual processing.
   Check answers Discussion Select an answer first
2. 7application · medium

   A bank deploys an AI system for credit approval. The model was trained on historical loan data that under-represented certain ethnic groups, raising concerns about disparate impact. The bank's risk team must identify and assess this risk. Which combination of actions best fulfills the risk identification and assessment phases?

   • AImmediately retrain the model with more diverse data without documenting the risk, to avoid regulatory scrutiny.
   • BAssign a data scientist to monitor the model's performance monthly and report any drift to the IT department.
   • CTransfer the risk by purchasing an insurance policy that covers discrimination lawsuits, without further analysis.
   • DConduct a bias audit using statistical metrics (e.g., disparate impact ratio) and document the potential compliance and reputational risks in the risk register.
   Reveal Solution Discussion Select an answer first
4. 8foundation · easy

   What is the primary purpose of risk reporting in an AI risk management framework?

   • ATo ensure that all AI risks are eliminated
   • BTo replace the need for risk monitoring
   • CTo provide stakeholders with timely and accurate information about AI risk exposure
   • DTo satisfy regulatory requirements only
   Reveal Solution Discussion Select an answer first
5. 9foundation · easy

   An organization is assessing the risk of an AI system that recommends medical treatments. The team assigns a probability of 0.2 that the system will give a harmful recommendation and estimates that each occurrence would cost $5 million in liability and corrective actions. Which type of risk assessment method is being used?

   • AScenario analysis
   • BQuantitative risk assessment
   • CQualitative risk assessment
   • DControl self-assessment
   Reveal Solution Discussion Select an answer first
6. 10application · medium

   A manufacturing company uses AI for predictive maintenance. The risk team is integrating AI risk management into the enterprise risk management (ERM) framework. Which action best supports this integration?

   • AUse only quantitative methods for AI risks, as they are more precise than qualitative methods used elsewhere.
   • BAlign AI risk categories with the existing ERM risk taxonomy and use the same risk scoring criteria as other enterprise risks.
   • CCreate a separate AI risk register that is not shared with the ERM team to avoid confusion.
   • DReport AI risks directly to the board, bypassing the ERM process, to ensure executive attention.
   Reveal Solution Discussion Select an answer first