Was it a pure coincidence that yesterday (August 25th, 2020) when my new book Artificial Intelligence for Audit, Forensic Accounting, and Valuation: A Strategic Perspective (Wiley) hit the market, PwC’s Bob Moritz committed to aggressively review and improve the quality of audits?*
The answer to the question is “yes” and I acknowledge it was a cheap tactic to get you to open the link – but if you have even slightest interest in revolutionizing and reinventing audit, just keep reading.
Why a new book on Artificial Intelligence in Audit?
For more than two decades now, I have been on both ends of audit – and neither gave me a sense of assurance and confidence. As an auditor, I felt pressured to complete things faster and with less resources. When being audited, I felt I was just a “check-the-box” item.
There was no objective reality There was no solid framework. There was no business value-added. If the universe of business problems is like a polluted ocean, legacy audit would be like a lonely castaway survivor dipping net in the water to catch fish – while simultaneously hoping to solve the pollution problem. In other words, too little, too lame, too limited, and too lost.
I felt that when it came to audit, we were simply playing roulette. The difference was that the casino and the gamblers, all seemed to be in a strange state of cohesion – a state often observed in softball, overly friendly audits.
Something needed to change even before the current crisis emerged (and this crisis is not Covid19).
The sudden and powerful rise of artificial intelligence became a gamechanger for business – a change for which audit is least prepared:
- It introduces new types of emergent risks for which audit firms are not ready.
- It forces audit firms to rethink their service delivery systems.
- It requires audit firms to stay a step ahead of the emergent risk vs. always playing the catchup
So, don’t be surprised to discover that Wirecard, the firm whose demise made Bob Moritz issue the call for change, offered a comprehensive service model for fraud detection. The ironies of our time!
If the existing pace of innovation in audit is a gauge for what lies ahead, we are in trouble. A new approach and a new book were necessary. As an Applied AI technologist, I took the lead to envision an audit automation model that can deliver constant, intelligent, and automated audit.
It is all about the emergent risk
AI has infused new risk in all aspects of business. Automation is the right answer to most business problems – and businesses have recognized that non-intelligent information technology can no longer provide a sustainable competitive advantage. Call it whatever – cognitive, AI, Fourth Industrial revolution, or intelligent automation – AI has now become a central part of our business.
The emergent risk comes from many sources – including from a company’s operations – where rapid automation is enabling a dynamic universe of interactive agents working in an adaptive and evolutionary configuration.
Decision-making by intelligent agents (human-machine or machine alone) requires constant for risk and bias.
New Audit Business Models with AI
I explain in my book that the audit will experience a surge in new business models, including:
Technology Provider: Audit firm will have the opportunity to develop, design, and install AI agents to help clients achieve better internal control and improve greater visibility to increase the understanding of inherent risks. These systems will be done for the internal use of the client firm and not for the external use. PwC’s cash.ai project is an example of that.
Audit as a Service: Audit firms can now deploy AI agents that perform continuous audit and track new developments in client’s business. These agents will be visible to the audit firm.
Audit the Auditor: Audit firms will acquire much better understanding of their audit teams and match the right people with clients.
Audit the Agents: As agents become more pervasive, governance of audit agents will become an important issue. Audit firms will need the ability to audit the agents used by the clients.
Assess the maturity: Audit firms will be able to help clients determine their maturity in terms of being able to audit their intelligent and non-intelligent digital workforce.
Niche Players: We will observe a sharp rise of audit-tech firms focused on niche areas of audit and these firms will sell their services to other audit firms.
Innovation is Key to Making Progress in Audit
We find ourselves in an extremely complicated position. Audit was not ready to face the world in the pre-AI-rise period – and now it seems even more incapacitated.
If audit needs to be relevant, this perpetual catchup of audit must end. Audit must grow up and face the reality. In my book I describe that as:
Audit has always been in a reactive mode and not a proactive mode. Scandal after scandal, and failure after failure, audit has functioned as a follower and never a leader of best practices. Being proactive means that audit must stay ahead of the emergent risk.
With great concern, I highlight something far more important: It is not enough to keep the pace, the pace of innovation in audit must exceed than that of the entities being audited.
The model presented is comprehensive
In my book I offer a model to show how the entire audit process should be reviewed and automated – with adaptive design elements in mind. It will require a blend of various technologies including machine learning (basic), deep learning, RPA, process mining, and other technologies.
It will also require focusing on the entire process. Hence, I try to synthesize various areas of innovation in my book to assemble a comprehensive model of intelligent automation (Figure 1).
This model has five areas for audit, including:
- Automated preaudit
- Automated risk assessment
- Automated Audit procedures
- Automated Audit reporting
- Automated post-audit management
This above model is developed and used by the American Institute of Artificial Intelligence (AIAI). AIAI is not approaching governance and audit problems as separate problems. And both problems have two dimensions to them:
- Using AI to perform audits
- Auditing AI system
The coexistence of these two dimensions leads us to keep our focus on the integration of the following areas:
- Audit standards for AI
- Audit automation
- ESG (Sustainability – Environment, Social, and Governance)
If we want to avoid a repeat of the Great Recession, or a total meltdown of the human civilization (which is quite possible as we have seen with Covid19), these three areas must develop simultaneously. Yes, it is about audit efficiency and quality – but more importantly it is about being responsible. Bob Moritz is right: a change is necessary – and AI will make it happen.
You can order my book on Amazon or Google. Also read my other article on ESG.
* See Financial Times article PwC pledges to review fraud detection after Wirecard scandal shakes industry by Tabby Kinder on August 25th, 2020