Abstract
Artificial intelligence (AI) technology has impacted businesses and industries as well as audit companies. With the emergence of AI-enhanced systems, many tasks performed by auditors can now be completed more efficiently by these technologies. Such systems are used in different audit tasks, such as risk assessment, audit planning, fraud detection, audit inquiry, transaction testing, inventory count, and document testing. AI platforms designed for auditing provide time-saving, higher efficiency and accuracy, minimized risks and biases, and improved audit quality. This chapter provides examples of AI platforms and tools developed by Big 4 audit firms and discusses the benefits and risks of implementing AI technology in auditing regarding the extant literature.
TopArtificial Intelligence Technology
More than 60 years ago, John McCarthy defined AI as “the science and engineering of making intelligent machines, especially intelligent computer programs”. In the 1950s and 1960s, people expected AI to develop rapidly into robots and computers that can mimic the intelligence of human (Bolander, 2019). Nevertheless, after the 1950s, much attention and scientific obscurity were not paid to AI for over a half-century. Following the developments in computer technology and the increasing rise of big data, AI gained more importance in the business environment and among the public (Haenlein & Kaplan, 2019).
AI supports businesses by automating the business processes, providing insights by analyzing data, and engaging with employees and customers (Hung & Sun, 2020). For this purpose, AI uses various techniques for simulating human cognition. The programs designed for specific purposes allow AI-based applications to drive cars, engage in dialogues, play chess or recognize skin diseases (Bolander, 2019). AI even affects every aspect of daily life through many applications such as personal assistants in mobile phones, cyber protection, and customization of products and services (PwC, 2017).
Although AI has some weaknesses compared to human intelligence, combining human experts and AI systems seems to provide a better solution. For instance, the Corti company in Denmark developed a pattern recognition algorithm that identifies probable cardiac arrest cases by listening to emergency calls. The algorithm had some false positives and negatives. Still, more cardiac arrests were detected, only at the cost of sending a few more ambulances than would have been sent otherwise (Bolander, 2019). Moreover, during the COVID-19 pandemic, AI-based applications helped small and medium-sized enterprises target customers online, make cash flow forecasts and facilitate HR activities for better operational strategies, and have better-informed financial planning through pricing and risk analysis. In this way, AI enabled companies to increase efficiency and reduce business risks by developing defense mechanisms and finding solutions for the challenges posed by the COVID-19 pandemic (Drydakis, 2022).
There are different subsets of AI technology, of which the most common ones are mentioned below.
Key Terms in this Chapter
Inherent Risk: Inherent risk can be defined as the probability of any material misstatement, independent of internal controls. So, inherent risk arises from the nature of business transactions or operations without any internal control implementation for mitigating risks.
Audit Inquiry: In addition to other audit procedures, audit inquiry is the process of gathering complementary information from the management, accountants, and knowledgeable persons inside or outside the company about the transactions, events, or anything that could assist auditors in evidence evaluation.
Control Risk: Control risk is the probability of any material misstatement due to the lack or malfunctioning of internal controls to detect or prevent errors and fraud.
Detection Risk: Detection risk is the likelihood that an auditor would not be able to identify material misstatements in the financial statements due to factors such as incorrect audit procedure application, incorrect audit testing methods, misinterpretation, or wrong assessment of audit results.
Natural Language Processing (NLP): NLP is an artificial intelligence field where computers read, analyze, understand, and interpret human language.