QuestionQ90

AI Governance and Risk

An IS auditor reviewing change-management procedures for an AI system finds inconsistent protocols for validating and verifying training data before model retraining. Which of the following represents the MOST significant risk in this situation?

Explanation

Training data validation and verification help prevent compromised, inaccurate, or poor-quality data from being used to retrain a model. If those controls are inconsistent, the retrained model can produce unreliable results and its operational reliability can degrade.

Community Discussion

No comments yet. Be the first to start the discussion!