A written, evidence-based verdict on whether an AI system is governed, safe, lawful and operating as advertised — produced against a named frame (ISO 42001, NIST AI RMF, DPDP) and delivered as a board-ready report.

What does an AI governance audit cover?

Policy and accountability, the 12-category risk register, per-system impact assessment, third-party AI register, evaluation regime, human oversight, and incident readiness.

How to use AI to support integrated ISO audits?

Use AI for evidence collection — Annex A gap analysis, artefact sampling, drafting statements of applicability — but never for sign-off. The lead auditor's judgement is what makes the report defensible.