Platformed

Which parts of an audit can AI automate today?

AI in audit · last updated 2026-09-02

AI can already automate substantial parts of evidence collection, document review, process documentation, risk identification, control assessment, testing preparation and report generation. It is strongest where the work is repeatable, evidence-led and governed by a defined methodology.

Examples include:

AI is less suited to decisions that depend heavily on commercial context, sensitive judgement or an auditor's accumulated understanding of management and the entity. It can prepare the ground for those decisions, but it should not disguise uncertainty or turn a judgement into an automatic answer.

The practical dividing line is simple: automate the legwork, not the accountability. If the system can make the evidence easier to understand and the work easier to review, it is useful. If it asks the auditor to trust an unexplained conclusion, it is not ready for audit work.

Platformed capability: Platformed prepares evidence collection, narratives, process maps, risks, controls, assessments, findings and reports. It does not sign the opinion or decide whether the resulting evidence is sufficient and appropriate; those decisions remain with the audit team.

References: IAASB automated tools and techniques resources · Platformed control assurance

← All 60 questions