AWS Certified AI Practitioner AIF-C01 Practice Question

A non-transparent, or black-box, model produces predictions without revealing a human-interpretable rationale for how specific input features lead to its output. Because its internal parameters and interactions are complex and opaque, users cannot easily trace or explain individual decisions. By contrast, models that provide readable if-then rules, rely on fully open data and code, or deliberately trade accuracy for interpretability are considered more transparent or white-box approaches.