A use-case audit evaluates an AI system against defined criteria: purpose, data, risks, classification level, control measures, documentation. It produces findings and corrective actions, and feeds the go-live decision.
Example
Review of a candidate-screening tool: data, bias tested, human oversight, documentation.