AI Policies, Procedures, and Organizational Training AAIR Practice Questions (Page 3)

Questions 11–15

1. 11expert · hard

   A large enterprise has an AI governance committee, but roles are unclear. The data science team develops models, the legal team reviews contracts, and the risk team conducts assessments, but no one owns the overall AI policy. A new AI model is deployed without a required risk assessment because the data science team thought legal was responsible. Which role assignment would most likely prevent this failure?

   • ACreate a shared responsibility model where all teams are jointly responsible for risk assessments, with no single owner.
   • BDesignate a single AI governance owner with authority to define roles, enforce the policy, and ensure risk assessments are completed before deployment.
   • CAssign the legal team the responsibility for all AI risk assessments and policy updates.
   • DAssign the data science team the responsibility for conducting risk assessments before deployment.
   Reveal Solution Discussion Select an answer first
2. 12foundation · easy

   What is the purpose of a training needs assessment?

   • ATo identify skill gaps across roles
   • BTo design training content
   • CTo evaluate training effectiveness
   • DTo deliver training sessions
   Reveal Solution Discussion Select an answer first
4. 13expert · hard

   A multinational corporation is integrating its new AI policy into an existing enterprise risk management (ERM) framework. The ERM framework uses a five-level risk severity scale, while the AI policy uses a three-tier classification. The audit committee wants AI risks to appear in the enterprise risk register with consistent severity ratings. Which integration approach is most effective?

   • AReplace the ERM five-level scale with the AI three-tier classification to simplify reporting.
   • BAssign all AI risks a 'high' severity in the ERM to ensure they receive attention.
   • CMap the AI three-tier classification to the ERM five-level scale using a defined translation matrix, and update the risk register accordingly.
   • DKeep the AI policy separate from the ERM framework and report AI risks only in the AI governance committee.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which metric is commonly used to evaluate AI training effectiveness?

   • AParticipant feedback scores
   • BNumber of trainers
   • CNumber of training sessions
   • DLength of training materials
   Reveal Solution Discussion Select an answer first
6. 15foundation · easy

   Why should AI training programs be tailored to different audiences?

   • ATo satisfy auditors
   • BTo make training longer
   • CTo reduce training costs
   • DBecause different roles have different learning needs
   Reveal Solution Discussion Select an answer first