Free ISACA AAIR Exam Dumps Questions & Answers

Which of the following is a risk practitioner ' s BEST recommendation to establish accountability for AI system outputs and decisions?

• A.  Formal documented role assignments with named owners
• B.  Continuous monitoring and key performance indicators (KPIs)
• C.  Regular reviews of resource allocation for AI projects
• D.  Centralized governance task force for model decision authority

In the context of AI risk mitigation, which of the following is the PRIMARY reason to have rollback capabilities in place for AI system updates?

• A.  To enable experimentation with new features
• B.  To remediate vulnerabilities introduced by coding errors
• C.  To correct unintended consequences
• D.  To facilitate more efficient analysis of inaccurate outputs

Which of the following MOST effectively reduces bias resulting from variances in free-form answers in the dataset?

• A.  Increasing the volume of available examples
• B.  Identifying and removing incomplete records
• C.  Scaling and standardizing inputs
• D.  Labeling unstructured data

Which of the following BEST enables an organization to comply with regulations on the use of biometric data by its AI models?

• A.  Establishing continuous monitoring of biometric capture system logs
• B.  Deploying liveness detection to recognize deepfakes and spoof attacks
• C.  Implementing strong governance to ensure biometric data is used only for its intended purposes
• D.  Performing vulnerability scanning of biometric endpoints and reporting results to regulators

A risk practitioner is evaluating training datasets for a new AI model. Which of the following approaches BEST reduces fairness risk during model development?

• A.  Selection of a single provider for large volumes of training data to reduce variance
• B.  Use of synthetic data to address training dataset imbalances
• C.  Representative sampling strategies combined with bias mitigation controls
• D.  Identification and resolution of data labeling inconsistencies