CORAA
Resources · AI in Internal Audit · India

AI in internal audit 2026.

Guides, copy-ready prompts, free working files and the standards, for internal audit teams, CA firms and finance heads in India. Start with the guide, pick your AI tool, then take the working file you need.

Read the 2026 guideOpen the prompt library
The short answer

What is AI doing in internal audit in 2026?

AI does two different jobs in internal audit in 2026. Chat assistants such as ChatGPT, Claude, Copilot in Excel and Gemini draft and structure: a risk and control matrix from a process narrative, walkthrough questions, a test step or a spreadsheet formula, an observation in five parts, a summary for the audit committee. Systems that apply fixed rules to every transaction in the books do the testing: duplicate payments, approvals outside limits, unusual journal entries, with each exception traced to its voucher. The first gives you a draft to verify. The second gives you a result you can run again and get the same answer.

What has not changed is who decides and who is responsible. Under Section 138 of the Companies Act, 2013 and Rule 13 of the Companies (Accounts) Rules, 2014, the audit committee or the Board, in consultation with the internal auditor, settles the scope, functioning, periodicity and methodology of internal audit. A chat assistant does not test a population and can invent a citation, so the internal auditor still owns the evidence, the rating and the conclusion. Personal data of employees, customers and vendors stays out of consumer AI tools unless you have authority to put it there.

Start here

Three things to open first

AI in Internal Audit 2026: the India guide
What AI does at each stage of an internal audit, what it cannot do, and what the internal auditor still owns. Read this first.
Open resource ->
AI in Internal Audit: October 2026 update
What changed recently and what it means for an internal audit function in India.
Open resource ->
Internal Audit AI Prompt Library
Copy-ready prompts by audit stage and process cycle, with an Excel download and an AI use log for the working paper.
Open resource ->
By AI tool

ChatGPT, Claude and Copilot in Excel for internal audit

One guide for each assistant an internal auditor is likely to have, plus two that apply to any of them: how to write a prompt that survives review, and how far a tool can go in writing the report.

ChatGPT for internal audit
Copy-ready prompts by audit stage, what not to paste, how to check the output and what to record.
Open resource ->
Claude for internal audit
How to set up a Project for one engagement with reference files, prompts and a worked flow.
Open resource ->
Copilot in Excel for audit testing
Eight audit tests with the plain-English ask and the classic Excel formula the answer should resemble.
Open resource ->
AI prompts for internal audit
Why most audit prompts fail, the parts of one that works, and worked prompts for RCM, walkthrough and observations. Works in any assistant.
Open resource ->
Can AI write an internal audit report?
A worked example from rough fieldwork notes to a five-part observation, with the review checklist a senior applies.
Open resource ->
AI agents

Agentic AI in internal audit and how to audit them

An agent takes steps by itself instead of answering one question at a time. These guides cover what that means for the audit team, and how to audit agents and AI systems the company itself runs.

Agentic AI in internal audit
What an AI agent can take over in internal audit, and what it should not.
Open resource ->
What is an AI agent?
A plain-words explainer for auditors and chartered accountants.
Open resource ->
How to audit AI agents
An internal audit programme for AI agents in the company: risks, controls, tests and evidence.
Open resource ->
ISO 42001 internal audit checklist
Auditing an AI management system, area by area.
Open resource ->
Do the work

Free tools and working files

A prompt gives you a draft. These hold the checked result: the RCM, the sampling plan, the observation, the follow-up, and the record of where AI was used.

Internal Audit AI Prompt Library
Pick the audit stage and the cycle, copy the prompt, and log the use.
Open resource ->
Internal Audit RCM Builder
Put the checked risk and control rows into an editable matrix with tests, evidence and reviewer columns.
Open resource ->
Observation Report Generator
Turn verified observations into a report pack with an executive summary and a management response tracker.
Open resource ->
Sampling Plan Generator
Record population, selection basis, sample size and the conclusion on deviations. The sample size stays your decision.
Open resource ->
ATR Tracker
Track owners, due dates, closure evidence and re-test results after the report is issued.
Open resource ->
AI Governance Internal Audit Workpaper
Audit the company’s own use of AI: inventory, ownership, data controls, changes, human review and incidents.
Open resource ->
Internal Audit AI Strategy Template
Approved uses, data boundaries, review gates and documentation rules, before AI use spreads informally.
Open resource ->
Standards and law

What AI leaves unchanged

ICAI lists its Compendium of Standards on Internal Audit (as on February 2026) as applicable from 1 April 2026. Check the ICAI Internal Audit Standards Board compendium before you cite a standard number in a report or an engagement letter.

Standards on Internal Audit (SIA)
The ICAI standards in one list, with plain-English summaries, what the internal auditor has to do and what to keep in the file.
Open resource ->
Internal Audit Applicability Checker
Test whether internal audit applies to the company under Section 138 and Rule 13.
Open resource ->
Internal audit in 2026: what has changed
What is different for an Indian internal audit function this year, including the two sets of SIA numbering.
Open resource ->
Appointment of internal auditor for FY 2026-27
Board resolution, the Section 138 rules and who can be appointed.
Open resource ->
For the finance head

What management should see every month

For the CFO, finance controller and audit committee member: cash, working capital, the month-end close and the exceptions worth an owner and a closure date. Testing every transaction for exceptions like these is what CORAA is built for; the files below work without it.

MIS Report Format Generator
A monthly MIS format in Excel: executive summary, P&L against budget and last year, working capital, cash flow and an action sheet.
Open resource ->
Month-End Close Checklist Generator
A day-by-day close calendar with task, owner, evidence and sign-off, for Tally, SAP or any ERP.
Open resource ->
Cash Conversion Cycle Calculator
Receivable, inventory and payable days, the cash tied up in working capital, and what a few days of improvement would release.
Open resource ->
Duplicate Payment Test Kit
The tests that find duplicate vendor payments in Tally or SAP data, the fields to pull, and an Excel register with a recovery tracker.
Open resource ->
Finance Controls Health Check
Yes, partly or no questions on payments, receivables, close, cash, payroll, compliance and access, with a score per area and a one-page board summary.
Open resource ->
Intercompany Reconciliation Template
The intercompany balance matrix, each mismatch with its likely cause, and draft elimination entries for group companies.
Open resource ->
SAP Audit Data Request Tables
The SAP tables and key fields to request for each process cycle, ECC or S/4HANA, with the Tally equivalent.
Open resource ->

Looking for checklists, RCMs and audit programmes by process cycle? They are in the internal audit resources hub.

FAQ

AI in internal audit questions

In two ways. Chat assistants such as ChatGPT, Claude, Copilot in Excel and Gemini draft and structure: an RCM from a process narrative, walkthrough questions, a test step or formula, an observation, a summary for the audit committee. Separately, systems that apply fixed rules to every transaction do the testing and trace each exception to its voucher. The first kind produces drafts to verify; only the second produces results you can test again and get the same answer.
Yes, for drafting and structuring, with your organisation’s authority and without personal data. They do not test a population, they can invent a section number or a threshold, and the internal auditor remains responsible for every fact and conclusion in the file. Verify the output against evidence before it enters a working paper.
No. AI takes over reading and drafting, and software can test every transaction instead of a sample. Deciding the scope with the audit committee, confirming facts with the process owner, judging cause and seriousness, and standing behind the report remain the internal auditor’s work.
Not without authority. Get the company’s approval for the tool and the class of data, check your plan’s data-use and retention settings, and keep personal data of employees, customers and vendors out of consumer tools. Most drafting needs only masked or described data, and a formula question needs only the column layout.
ICAI lists its Compendium of Standards on Internal Audit (as on February 2026) as applicable from 1 April 2026, and that set includes a standard titled Use of Tools. We have not verified its text, and ICAI’s page does not say whether the set is mandatory or recommendatory, so check the ICAI Internal Audit Standards Board compendium before you cite it. The duties on evidence, documentation, review and reporting apply whether or not a tool helped.