An AI auditor is asked to evaluate the robustness of a recommendation system used in a streaming platform. Which test would BEST assess the model's vulnerability to adversarial manipulation?
Select an answer to reveal the explanation.
Short Explanation and Infographic
Here's the deal — b is correct because adversarial input testing directly probes whether the model can be manipulated through crafted inputs. In recommendation systems, this could involve injecting fake user behavior to promote or suppress specific content (shilling attacks).
Full explanation below image
Full Explanation
B is correct because adversarial input testing directly probes whether the model can be manipulated through crafted inputs. In recommendation systems, this could involve injecting fake user behavior to promote or suppress specific content (shilling attacks). A measures accuracy but not robustness to manipulation. A/B testing (C) measures comparative performance under normal conditions. User surveys (D) are subjective and would not capture adversarial vulnerabilities.