AI Audit Readiness Checklist Prompt
Prompt
You are a Controller preparing for an audit in an environment where the finance team uses AI tools. Current AI usage: [Paste: tools in use, processes assisted by AI, frequency] Build an audit readiness checklist: Documentation: - AI usage policy (current version, approved by management) - List of approved AI tools and use cases - Data classification policy (what can/cannot go into AI) - Prompt library (approved prompts, version controlled) Process controls: - Human review evidence for every AI-generated output used in financial reporting - Segregation between AI operator and reviewer - Exception log (instances where AI output was rejected or significantly modified) Audit trail: - Log of AI tool usage (who, when, what prompt, what output) - Input data preserved (can recreate what was fed to AI) - Output data preserved (what AI produced before human edits) - Final version preserved (what was actually used/published) Testing: - Periodic accuracy testing results (prompt validation) - Comparison of AI outputs to manually-prepared versions - Error rate tracking over time Format: Audit readiness checklist with status column (in place / in progress / gap). Include suggested evidence for each item.
Why it works
Auditors are starting to ask about AI usage. Being prepared with documentation and controls demonstrates responsible adoption.
Watch out for
Risks: Lack of AI documentation could result in scope expansion or adverse findings. Get ahead of auditor questions. Control: Internal audit reviews AI readiness before external audit begins.
Used by
Finance Teams