A financial audit allows you to make a detailed assessment of the financial accounts of a company, and the largest companies are required to do one every year. AI, with its great computing power, can be a very useful tool for the auditor’s job. The Master CARF (Control Audit and Financial Reporting) of the University of Paris Dauphine-PSL and the Business Risk Services team of Grant Thornton have joined forces to conduct the study on “The place of artificial intelligence in audit practices.”

As companies face ever-increasing compliance requirements, auditors’ assignments are becoming increasingly diversified, whether they are internal or external.

The survey

The survey was conducted over the past year: 120 internal and external auditors responded to the survey, 2/3 of which were internal:

• Almost 90% of them work in large or very large companies (53%);
• Half of the companies are listed;
• The main sectors of activity represented are Industry, Banking and Services.

The vast majority of these 120 auditors had more than 3 years of seniority in the function and the majority of the 120 came from Masters in management and auditing (CCA or others), followed by business schools (BAC + 5).

A widespread but classic use of AI

The most widespread AI tools are the following:

• Data visualization tools: 2/3 of the respondents use data visualization tools, 49% use
  Data Visualization tools: 2/3 of respondents use data visualization tools, 49% of them frequently;
• Data analysis tools “embedding” generic controls pre-configured by the editor: half of the respondents already use these tools;
• Data analysis tools that allow you to parameterize controls and develop your own scripts: almost half of the respondents parameterize.

Only 5% of companies use drones, often for inventories, 16% use robots and 40% use machine learning.

Summary of the use of AI tools for audit practices

Nearly 60% of auditors thus reported using at least one of the AI-based tools while 13% never used any. The simplest ones are the most used: while 50% use data analysis tools, only 3% use bots on a regular basis, 11% use machine learning technologies and 13% use visual recognition tools.

These AI tools are primarily used to:

• The control of the Accounting Entry File;
• Data reconciliation;
• Search for anomalies or atypicals;
• Sampling.

The main barriers to the development and use of AI-related tools in auditing

For 63% of respondents, the main obstacles are financial and technical.

The technical obstacles are important:

• 73% of respondents believe that the use of these tools requires a significant investment of time to
  73% of respondents believe that the use of these tools requires a significant investment of time in order to fully understand the company’s data architecture and the operation of computerized processes;
• 66% that the lack of IS skills among auditors is an obstacle;
• 65% that auditors’ lack of knowledge about the possibilities offered by data is a hindrance;
• 69% that access to data is problematic.

More efficient audits

When asked about the opportunities brought by AI, 84% felt it brought an increase in the reliability of work and conclusions, including:

• Improving the efficiency of tests for 94% of auditors;
• Optimizing test sampling for 91%;
• For risk coverage for 89%;
• For the relevance of the audit plan and risk identification for 79%;
• AI tools facilitate auditing and increase the scope of work that can be done remotely for 65%;
• They enable continuous audit development for 78%.

Regarding the writing of audit reports, only external auditors seem to see it as an opportunity.

The auditor of tomorrow will not be a robot

Respondents to the survey believe that their job will not be disrupted by the development of AI. For 89% of them, the auditor of tomorrow will have to have a serious culture of data analysis and 96% are certain that it will not be a robot!

Translated from Quelle est la place de l’intelligence artificielle dans les pratiques d’audits financiers ?