Data Management Controls AAISM Practice Questions (Page 3)

Questions 11–15

1. 11application · medium

   A research institute develops an AI model using public datasets and proprietary data. The institute wants to ensure that the model's outputs can be audited by external reviewers. What should be implemented to support this?

   • ADocument the provenance of each dataset and the transformations applied.
   • BImplement a data retention policy.
   • CUse a data classification scheme for all datasets.
   • DStore all data in a single repository.
   Reveal Solution Discussion Select an answer first
2. 12application · medium

   A marketing firm uses customer data to train a churn prediction model. The data includes email addresses, purchase history, and demographic information. The firm wants to reduce the risk of re-identification if the data is breached, while still maintaining the model's predictive accuracy. Which privacy-enhancing technique should be applied?

   • AUse only aggregated data for training.
   • BEncrypt the dataset with a strong encryption algorithm.
   • CApply k-anonymity to the dataset before training.
   • DRemove all demographic information from the dataset.
   Reveal Solution Discussion Select an answer first
4. 13expert · hard

   A manufacturing company uses AI to predict equipment failure. The model is trained on sensor data that is collected in real time. The company wants to ensure that the data used for training is of high quality and that the model remains accurate over time. Which combination of lifecycle controls is most appropriate?

   • ADispose of sensor data immediately after use.
   • BRetain all sensor data indefinitely for future retraining.
   • CImplement data validation at acquisition and periodic model retraining with fresh data.
   • DUse only historical data for training and avoid real-time data.
   Reveal Solution Discussion Select an answer first
5. 14foundation · easy

   Which technique is commonly used to protect individual privacy in AI training datasets by adding controlled noise?

   • AEncryption
   • BHashing
   • CDifferential privacy
   • DTokenization
   Reveal Solution Discussion Select an answer first
6. 15expert · hard

   A large enterprise is adopting AI across multiple departments. Each department has its own data governance practices, leading to inconsistent data classification and access controls. The enterprise wants to implement a unified governance framework without disrupting ongoing AI projects. Which approach is most effective?

   • AEstablish a governance committee to develop standards and roll them out incrementally.
   • BAllow each department to continue its own practices but require them to report to a central authority.
   • COutsource data governance to a third-party vendor.
   • DImmediately enforce a new enterprise-wide data classification standard across all departments.
   Reveal Solution Discussion Select an answer first