THE FUTURE OF AUDIT
By Accountancy Students CALEB CASTRO, LAM YOU KANG and NG YEN LYN, as well as Information Systems students JASON KOH AND AMOS TAN
This article is an adaptation of a project report submitted by the writers for the Accounting Analytics Capstone Course at the SMU School of Accountancy (SoA). Supervised by Associate Professor Gary Pan and Professor Venky Shankararaman, it is part of SMU-X – a learning pedagogy that allows students to apply an interdisciplinary approach as they develop innovative solutions for real client projects with mentorship by faculty and the industry partner.
For this project, the students, as aspiring young accountants and auditors, envisioned how the future of audit, and the profession, will change as the traditional roles are transformed with the help of new technologies, to become more efficient, strategic and catalytic. They focused on how accounting processes can be simplified and streamlined, and how applied data-driven analytics and insights can deliver future solutions to current real-world challenges.
This project is a collaboration between Deloitte Singapore and SMU SoA to develop future-ready audit professionals.
AN AI PERSPECTIVE
Artificial Intelligence, or AI, is fast changing the way we do work in the audit profession. In this article, we posit a vision of what the future of the profession will be like with the rise of AI, and examine how AI technologies can potentially be integrated into the future audit process.
Before we are able to look into the future of audit, we must first define the future client. From experience, it is almost always that the advancement of the audit methodology is limited by how technologically advanced the client is. In order for us to proceed with our exploration into the future without restrictions, we must presume that our future client are Industry 4.0 organisations.
In the Industry 4.0 world, where companies are completely digitalised, employ robotics and the Internet of Things, and allow for the creation of a mirror world, audit firms will need to integrate AI technologies into the way audit is performed in order to stay relevant to their clients. We call this next step in the evolution of audit “AI-assisted audit”.
In this future world, the future audit firm transforms into a technology company which specialises in audit. Unlike how audit firms operate today, the future audit firm operates primarily over the web, as being physically on-site to conduct audits has become redundant. Instead, the future audit firm provides audit as an online service comprising three modules – control, data and reporting.
THE CONTROL MODULE
The control module generates test cases through the use of a self-coding AI, and tests them against a company’s controls which are digitised and defined within an enterprise system. The AI generates syntax using a technique known as inductive programme synthesis, where it is first given a set of input(s) and output(s), after which, the AI obtains codes from existing programmes and intelligently induces the programmes.
By using this self-coding AI, the control module is able to smartly adapt and provide testing on controls for other clients with other requirements. Furthermore, as an automated process, it allows controls to be continually tested even if a company introduces new controls. This function significantly changes the role of the auditor from that of manually testing the controls to that of checking the syntax generated by the AI to ensure that controls are tested fairly.
THE DATA MODULE
The data module checks the reliability of the clients’ data by leveraging on Big Data. Advanced text mining techniques are used to extract structured and unstructured data from the Internet and various data sources to identify fresh insights. Financial analytics and AI firm Kensho is a prime example of this – it is able to process large amounts of data in real time from structured sources such as financial databases, as well as unstructured data from the news, political reports, social media, etc. Auditors are able to harness the power of Big Data to perform reasonableness tests and substantive analytical procedures to determine whether a company’s forecasts, valuations and transactions are conducted in a true and fair manner.
THE REPORTING MODULE
The reporting module focuses on generating hypotheses to understand why and how anomalies occur. Audited financial statements may be generated at different time periods such as real time, daily, weekly, monthly and quarterly. Machine learning algorithms such as generative design, counterfactual regret minimisation, and reinforcement learning are used to generate possible reasons as to why and how fraud or misstatements occur. By designing a set of constraints – the company’s internal controls, codified accounting standards, and the objective to explain the anomaly – the module replicates the anomaly by running through every possibility of how the anomaly might occur given the constraints. The hypotheses generated will then be statistically analysed and ranked by likelihood of it causing the error. This allows auditors to judge which scenarios are most likely, and subsequently investigate the more pertinent ones. This streamlines the investigation process and results in effort and time saved for auditors.
To help explain how these modules and their AI technologies are applied, please refer to an example of an AI-assisted audit on a purchasing order (Table 1).
When Industry 4.0 and, consequently, AI-assisted audit become a reality, there will be changes to the structure of the audit firm. One example is manpower. AI will completely replace the staff headcount currently needed to carry out tasks involving control, data and reporting. This means that the job scope and skill sets required from this group of employees will be different – they will need to have the technical knowledge to maintain, modify and update the AI technologies.
Transiting from traditional audit to AI-assisted audit may take time. It requires technological evolution and is dependent on the clients’ level of technological adoption as well as approval from standard-setters and regulators. Nevertheless, with the impressive progress of AI in many fields, the audit profession ought to actively embrace its potential, rather than shy away from the uncertainties.
WHAT EXPERTS SAY
“Technology is changing, and will continue to change, the accounting profession. The advent of technology means that mundane and tedious work in the accounting profession will be automated. However, accounting professionals will play a more important role in processing and generating information that is critical for corporate decisions and financial reporting. Such changes suggest that accounting professionals should develop skills that are required in the future – tech-savvy, critical thinking skills and the ability to make sound professional judgement.” Lee Kong Chian Chair PROFESSOR CHENG QIANG, Dean, SMU School of Accountancy
“The world of audit has evolved over the past two decades and continues to evolve, with stronger regulatory demands, increasing data volumes and exponential growth in technological capacity. The introduction of new technologies and platforms, the rise of e-commerce, and the shift from manual to automated processes in organisations are also major factors in this evolution.
At Deloitte, we believe that the audit of the future is all about innovation. In fact, this future is already upon us – we are seeing exponential technologies and emerging operating models that are leading the way in transforming traditional audit processes. These include Artificial Intelligence and Cognitive Computing, Robotics, Cloud Computing, Internet of Things, Blockchain and Crowdsourcing.
There are significant benefits – speed, cost efficiency, better data access and improved audit quality, and to reap these benefits, audit professionals will need to be well versed in the new technologies and data analytics.
These are exciting times for the profession as it looks to reinvigorate and reinvent itself, and it is encouraging that the academia has started incorporating new modules as part of their curricula to equip future audit professionals with the skills and knowledge needed to deliver the audit of the future. It is equally important to have the standard-setters and regulators coming onboard this journey early, to provide guidance to the audit practitioners.” SANJAY PANJABI, Audit & Assurance Innovation & Analytics Leader, Deloitte Singapore
This is an adaptation of a project report submitted by the writers for the Accounting Analytics Capstone Course at the SMU School of Accountancy.
This article was first published in the IS Chartered Accountant Journal, April 2018. Reproduced with permission from the Institute of Singapore Chartered Accountants.
Last updated on 13 Apr 2018 .