ISACA Certified Artificial Intelligence Auditor (CAIA) — Question 48
Which of the following is the GREATEST data quality risk when using an AI tool to assist with audit procedures?
Answer options
- A. Utilizing unstructured data sources without standardized preprocessing
- B. Training models based on historical audit results generated prior to AI adoption
- C. Embedding AI audit tools in transactional systems without user training
- D. Applying automated anomaly detection without human oversight
Correct answer: A
Explanation
The greatest risk comes from utilizing unstructured data sources without standardized preprocessing, as this can lead to inconsistencies and inaccuracies. Training models on historical data prior to AI adoption (B) is less risky because it is based on established results. Embedding AI tools without user training (C) and applying automated anomaly detection without human oversight (D) also pose risks, but they do not directly affect data quality as severely as using unstructured data does.