AAIA Practice Test

AAIA Practice Test practice questions, page 15 of 23, with answer review and explanations.

Question 71

An audit reveals inconsistencies in the retention period of data used for AI training. What is the BEST recommendation to mitigate this risk?

  1. Enhance licensing agreements to include retention clauses.
  2. Establish a dedicated data clustering team for ongoing oversight.
  3. Implement automated data cleansing tools to enforce retention policies.
  4. Assign data ownership roles to the AI model development team.

Question 72

An IS auditor reviewing an AI-based data governance program discovers that there is no formal data classification policy in place. Which of the following is the MOST likely consequence of this gap?

  1. Inappropriate handling of sensitive data during AI model training
  2. Delayed AI model deployment due to data inefficiencies
  3. Reduced data cleansing accuracy due to lack of categorization
  4. Unauthorized data sharing with contractors

Question 73

To ensure stakeholders are adequately prepared to address workforce impacts, which of the following is an IS auditor’s BEST recommendation for an organization’s AI incident response plan?

  1. Review the frequency of AI security incidents and updates to the response plan.
  2. Verify that the organization has an AI ethics committee to discuss potential workforce impacts.
  3. Validate that regular cross-functional AI incident drills and skill gap assessments are performed.
  4. Confirm that all employees have completed a standardized online AI awareness course within the last year.

Question 74

An organization seeks to sustain effective AI governance and risk management amid rapidly evolving AI technologies. Which of the following represents the MOST effective course of action?

  1. Provide role-specific AI training to technical staff.
  2. Outsource AI training to external vendors.
  3. Conduct comprehensive AI training for senior management.
  4. Integrate continuous AI training into security awareness programs

Question 75

An AI tool is being implemented for a regional healthcare organization. Which of the following training methods BEST ensures the AI output does not reveal whether someone's personal data was used?

  1. Supervised learning with labeled patient records
  2. Data augmentation during training to improve privacy
  3. Differential privacy applied during model training
  4. Transfer learning using public health data sets