Glossary
Updated
A rigorous, often independent evaluation of an AI system’s data, code, outcomes, and governance processes (covering bias checks, privacy controls, performance validation, and regulatory adherence), culminating in a formal audit report.
Example
A healthcare provider brings in an external audit firm to review its diagnostic AI tool. Auditors examine patient-data diversity, run the model on withheld test cases representing different demographics, verify that all patient records comply with privacy rules, and issue a comprehensive report with findings and remediation steps, ensuring transparency and regulatory compliance.