Which of the following is the MOST effective approach for an organization to reduce the impact of adversarial attacks on AI models?
Answer(s): A
Adversarial testing exposes the model to deliberately manipulated inputs so weaknesses can be identified before attackers exploit them. Regular retraining with representative adversarial examples strengthens model robustness and helps maintain resilience as attack techniques evolve.
Which of the following provides the GREATEST assurance that an AI model will not disclose if a specific individual’s information was used to train the model?
Differential privacy adds carefully controlled statistical noise so that the presence or absence of any single individual’s data has minimal effect on the model’s outputs. This provides strong protection against attempts to infer whether a specific person’s information was included in the training dataset.
Which of the following is the MOST important reason to implement human-in-the-loop for AI tools that work with large, complex datasets?
Answer(s): B
Human-in-the-loop oversight provides a critical control for reviewing AI-generated outputs before they drive consequential decisions. This helps detect and correct biased, inaccurate, or inappropriate results, thereby reducing ethical and governance risk when AI operates on large and complex datasets.
A retail business has partnered with a technology service provider to outsource AI-based customer behavioral analysis.Which of the following is MOST important for an IS auditor to review?
Answer(s): D
Applying need-to-know principles limits the customer transaction data shared with the AI service provider to only what is necessary for the outsourced analysis. This reduces privacy, confidentiality, and third-party exposure while supporting data minimization and appropriate access control.
Which of the following is the MOST important reason for organizations to consider social risk associated with the output generated by AI models during the design phase of an AI system?
AI-generated outputs can materially influence human decisions, perceptions, and behavior at scale. Considering this social impact during system design helps the organization identify and mitigate risks such as harmful persuasion, misinformation, manipulation, or other unintended societal consequences before deployment.
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.