Change Management Specific to AI AAIA Practice Questions (Page 1)

Questions 1–5

1. 1expert · hard

   A large enterprise is transitioning from traditional IT change management to AI-specific change management. They have a mature ITIL-based change process. The AI models are used in critical customer-facing applications. What is the most important difference they must address?

   • AAI changes require continuous monitoring and validation because model behavior can drift over time
   • BAI changes are identical to traditional IT changes and require no special process
   • CAI changes do not require approval from the change advisory board
   • DAI changes are always reversible with a simple rollback
   Reveal Solution Discussion Select an answer first
2. 2foundation · easy

   Why is communication important when implementing an AI change?

   • ATo hide the change from users until it is fully deployed
   • BTo avoid the need for training
   • CTo ensure stakeholders are aware of the change and its potential impact
   • DTo speed up the deployment process
   Reveal Solution Discussion Select an answer first
4. 3foundation · easy

   At which stage of an AI model's lifecycle do changes to the training data typically occur?

   • AOnly during initial model development
   • BOnly during deployment
   • COnly during monitoring after deployment
   • DDuring data preparation, model training, and retraining
   Reveal Solution Discussion Select an answer first
5. 4foundation · easy

   What does the process of managing AI model updates in production typically involve?

   • ATesting the updated model, obtaining approval, and deploying it with monitoring
   • BDeleting the old model and immediately using the new one
   • COnly updating the model code, not the data
   • DReplacing the model without any testing or approval
   Reveal Solution Discussion Select an answer first
6. 5expert · hard

   A company has a predictive maintenance model for industrial equipment. The model's accuracy has declined over the past month. The operations team wants to retrain the model immediately, but the data science team is concerned about the cost and time. What is the best approach?

   • ARetrain the model immediately because accuracy is declining
   • BDeploy a new model without any testing
   • CIgnore the decline and continue using the current model
   • DConduct a cost-benefit analysis and a risk assessment before deciding to retrain
   Reveal Solution Discussion Select an answer first