ISACA Certified Artificial Intelligence Auditor (CAIA) — Question 25

A healthcare organization uses an AI model to analyze patient data and provide diagnostic recommendations. Which of the following MOST effectively detects data drift related to the model’s predictions?

Answer options

• A. Using adversarial testing to simulate scenarios that stress test the model’s predictions
• B. Conducting periodic model retraining to ensure alignment with updated patient data
• C. Comparing incoming patient data distributions with the training data set
• D. Applying overrides to allow healthcare professionals to correct the AI model's recommendations

Correct answer: C

Explanation

The correct answer, C, involves comparing incoming patient data distributions with the original training dataset, which is essential for detecting data drift. Option A focuses on testing the model under stress but does not directly address data drift. Option B, while important for keeping the model updated, does not specifically identify drift. Option D allows for corrections but does not effectively detect changes in data patterns.