Fororganization-wide adoptionof AI, the MOST important change management consideration is that the organization hassuitable data governance and infrastructure readiness(D). Without robust data governance, AI systems may rely on poor-quality, noncompliant, or insecure data; without appropriate infrastructure (compute, storage, monitoring, integration), implementations will be fragile, unreliable, and risky. AAIA stresses thatAI readinessincludes governance structures, data stewardship, and technical foundations.
Senior leadership involvement (A) is critical for sponsorship and culture, but even strong leadership cannot compensate for fundamentally inadequate data governance and infrastructure. Option B (shorter training cycles) focuses on user adoption but not core risk and control issues. Option C (phased implementation and stage gates) is good project discipline, but still depends on the underlying readiness. Therefore, from an AI risk and governance perspective, ensuringdata governance and infrastructure readinessis the primary prerequisite for successful, safe change management in AI adoption.
References:
ISACA,AAIA Exam Content Outline- Domain 1 and Domain 2: Governance of AI; AI Operations and Infrastructure.
ISACA guidance on AI readiness assessments and data governance as a precondition for AI transformation.

The confusion matrix is a key performance evaluation tool in machine learning and AI auditing. According to the AAIA™ Study Guide, a confusion matrix presents detailed information about actual versus predicted classifications, allowing auditors to assess accuracy, precision, recall, and F1 scores.
"A confusion matrix reveals not just how often predictions are correct, but also the types of errors being made-false positives and false negatives-thereby providing a clear view of the model's predictive reliability." Adversarial testing evaluates robustness, group analysis identifies bias across subgroups, and latency testing examines performance speed-not predictive accuracy. Thus, D is the most relevant for ensuring correct predictions.
Reference: ISACA Advanced in AI Audit™ (AAIA™) Study Guide, Section: "AI Operations and Performance," Subsection: "Model Evaluation Metrics"

The F1 score is the harmonic mean of precision and recall, offering a balanced measure of model accuracy, especially when there is an imbalance in the classes (e.g., more "no-maintenance" than "needs-maintenance" outcomes). According to the AAIA™ Study Guide, the F1 score is used to evaluate how well the model identifies true positives while balancing the risk of false positives and false negatives.
"An F1 score summarizes the model's ability to correctly identify relevant events-in this case, true maintenance needs. A 76% F1 score means the model is relatively balanced and effective at catching maintenance requirements without generating too many false alerts." Thus, D is correct. Option A misrepresents recall as predictive accuracy. Option B misinterprets accuracy.
Option C has no direct basis in the excerpt.
Reference: ISACA Advanced in AI Audit™ (AAIA™) Study Guide, Section: "AI Operations and Performance," Subsection: "Understanding Evaluation Metrics and Model Cards"

Ethical risk begins with the nature of the data being used. The sensitivity and origin of training data (option B) determine whether the model risks violating privacy, perpetuating historical bias, or using data that was collected without proper consent.
AAIA identifies data provenance, data sensitivity, and lawfulness of processing as central to ethical AI.
Training data that contains sensitive attributes (e.g., health, ethnicity, gender, financial status) must be reviewed carefully for compliance and ethical impacts.
Option A (diverse outputs) relates to performance, not ethics. Option C (update frequency) is part of lifecycle management, not ethical sourcing. Option D (cleaning methods) ensures quality but does not address ethical risk if the underlying data is inappropriate or unlawfully sourced.
Thus, sensitivity and origin of training data are the primary ethical factors.
References:
AAIA Domain 5: Ethical Principles, Fairness, Data Sensitivity Considerations.
AAIA Domain 1: Privacy and Data Governance Programs.

Assessing data lineage provides insight into the origin, flow, and transformation of data across its lifecycle, which is crucial for validating data governance. The AAIA™ Study Guide states that data lineage is essential to ensure the accuracy, consistency, and trustworthiness of data used in training AI models.
"Traceability of data sources is a core tenet of effective data governance. Data lineage validation ensures data quality, prevents unauthorized modifications, and maintains auditability." BIA (A) focuses on impact, not data quality. Reviewing SDLC (B) is broad and may not highlight data- specific risks. Penetration testing (C) addresses security, not governance. Therefore, D is the best method.
Reference: ISACA Advanced in AI Audit™ (AAIA™) Study Guide, Section: "AI Governance and Risk Management," Subsection: "Data Quality, Integrity, and Governance Practices"