Change Management Specific to AI AAIA Practice Questions (Page 4)

Questions 16–20

1. 16foundation · easy

   In the context of AI systems, what does change management primarily aim to ensure?

   • AThat changes to AI systems are controlled, assessed, and documented to maintain reliability and compliance.
   • BThat changes are automatically rolled back if performance drops by any amount.
   • CThat only the model code is versioned, while data changes are ignored.
   • DThat all changes are implemented as quickly as possible to keep the system current.
   Reveal Solution Discussion Select an answer first
2. 17expert · hard

   A multinational company is deploying an AI HR system that screens job applicants. The system is being updated to include new criteria. The change must comply with various regional regulations. What is the most important governance consideration?

   • ADeploy the change in all regions simultaneously to ensure consistency
   • BEnsure the change is reviewed by a cross-functional team that includes legal, HR, and data science
   • COnly have the data science team approve the change
   • DAvoid documenting the change to protect intellectual property
   Reveal Solution Discussion Select an answer first
4. 18foundation · easy

   Which of the following is a key element of governance procedures for AI changes?

   • AAllowing any team member to approve changes
   • BSkipping testing if the change is small
   • CDocumenting the change and its approval
   • DKeeping the change a secret until after deployment
   Reveal Solution Discussion Select an answer first
5. 19foundation · easy

   What is the primary purpose of a rollback plan for an AI change?

   • ATo delete all training data
   • BTo speed up the deployment process
   • CTo revert to the previous stable version if the change causes issues
   • DTo automatically improve the model's accuracy
   Reveal Solution Discussion Select an answer first
6. 20application · medium

   A credit scoring model's performance has degraded because of economic changes. The data science team proposes retraining the model with recent data. What should they do before deploying the retrained model?

   • AImmediately deploy the model to production to minimize the impact of the degraded performance
   • BRoll back to the previous model version
   • CNotify all customers that their credit scores may change
   • DValidate the retrained model on a holdout dataset that reflects current economic conditions
   Reveal Solution Discussion Select an answer first