Isaca Updated AAIA Exam Questions and Answers by laurie

An AI model inventory documents the models in use, their purposes, and how they support specific business functions. According to the AAIA™ Study Guide, maintaining a comprehensive AI model inventory allows auditors to trace model objectives, performance metrics, and use cases back to business goals.

“A well-maintained AI model inventory supports governance and alignment by offering a centralized view of model functions, business integration, and ownership. It ensures transparency and strategic coherence.”

While policies and assessments are important, only the inventory directly shows which AI models exist and their connection to organizational objectives.

When organizations leverage off-the-shelf AI models, effective vendor management is critical to ensure operational reliability, compliance, and long-term support. The ISACA Advanced in AI Audit™ (AAIA™) Study Guide highlights that the " establishment of clear contractual terms regarding responsibilities for ongoing model updates, maintenance, support, and incident response is essential for managing third-party AI risks. "

By clearly defining the roles and expectations for updates and support (option B), organizations reduce the risk of unaddressed vulnerabilities, outdated models, or unclear recourse in the event of an incident or system failure. This approach supports ongoing risk management and ensures that both parties understand their obligations throughout the model’s lifecycle.

While market research, vendor accreditation, and contract review by information security are important due diligence steps, they do not directly address the need for clarity in ongoing vendor responsibilities, which is critical for effective governance and sustained operation of AI solutions.

Ensuring that AI tools are trained on properly licensed and documented data sets is critical to avoiding copyright infringement and legal exposure. The AAIA™ Study Guide emphasizes using platforms with certified and traceable training data to meet ethical and legal standards.

“Organizations must verify the provenance and licensing of data used to train AI systems. Platforms that certify data sources reduce the risk of using protected intellectual property without consent.”

Manual review (A) is resource-intensive and may not detect embedded copyright violations. Labeling (C) is not sufficient for legal protection. Suspension (D) may be excessive without first attempting remediation. Thus, B is the most strategic and effective recommendation.

According to the ISACA AAIA™ Study Guide, facial recognition systems are prone to significant ethical risks related to fairness and bias. These systems often exhibit higher error rates for specific demographic groups—particularly based on race, gender, and age—due to non-representative training datasets. For an IS auditor, this represents a major compliance and reputational risk, as biased outcomes can lead to discriminatory access or false accusations. While data security (Option A) and the lack of standards (Option C) are relevant, the fundamental ethical imperative in AI governance is ensuring that biometric applications do not disproportionately disadvantage protected groups.