IAPP Artificial Intelligence Governance Professional (AIGP) — Question 57

All of the following types of testing can help evaluate the performance of a responsible AI system EXCEPT:

Answer options

• A. Risk probability/severity.
• B. Adversarial robustness.
• C. Statistical sampling.
• D. Decision analysis.

Correct answer: C

Explanation

Statistical sampling focuses on selecting a representative subset from a larger dataset, which does not directly measure the performance of an AI system. In contrast, the other options—Risk probability/severity, Adversarial robustness, and Decision analysis—are directly related to evaluating and ensuring the functionality and reliability of responsible AI systems.