An IS auditor reviewing an AI-driven HR screening model for bias discovers the system favors candidates of a specific gender for a technical position, despite candidates showing similar qualifications.Which of the following is the MOST likely source of the bias?
Answer(s): B
Training data is the most likely source because historical hiring data may contain existing gender imbalances or discriminatory patterns. When a model learns from such data, it can reproduce those patterns even when candidates have comparable qualifications.
An IS auditor is assigned to audit a key area of an organization’s data governance framework, including the adequacy of data protection measures against data poisoning and AI attacks.Which of the following areas is the auditor’s PRIMARY consideration?
Answer(s): A
Data security is the primary consideration because protection against data poisoning and AI-specific attacks depends on controls that preserve the confidentiality, integrity, and availability of data throughout its lifecycle. Effective security measures help prevent unauthorized manipulation of training, validation, and operational data.
Which of the following is the PRIMARY benefit of a robust data governance framework for an organization that uses AI extensively?
Answer(s): C
A robust data governance framework establishes accountability, policies, controls, and oversight for how data is managed across its lifecycle. Its primary benefit is helping ensure data-related AI risks are identified, controlled, and maintained within the organization’s approved risk tolerance.
An organization’s AI governance committee primarily relies on a key performance indicator (KPI) for the average confidence score of its customer service chatbot’s responses.Which of the following is the MOST significant limitation of this metric for governing the chatbot’s performance?
An average confidence score reflects how certain the model is about its responses, but it does not establish whether those responses are correct, useful, or aligned with business objectives. A chatbot can produce highly confident but incorrect answers, so relying on this metric alone can obscure poor user outcomes and material business impact.
When utilizing a machine learning (ML) model to predict whether a wind turbine electricity generator will fail, which model evaluation metric should be the PRIMARY focus?
The primary focus should be on Recall, as it measures the model’s ability to correctly identify actual failures. In critical applications like wind turbine failure prediction, missing a true failure (false negative) can result in high costs or safety risks, making recall the most important metric.
Post your Comments and Discuss ISACA AAIA exam prep with other Community members:
Our website is free, but we have to fight against AI bots and content theft. We're sorry for the inconvenience caused by these security measures. You can access the rest of the AAIA content, but please register or login to continue.