Audit from Spy to A.I.


Audit was one of the oldest professions in the world, some archeologist found them in many cultures like the Egyptian and Greeks. Back then their job was to verify someone else’s report, and in ancient times they don’t have paper so they must listen to the report and verify it later. The first well-documented one was the spies of king Darius, that acted as the “king’s ears” by watching a province of the kingdom and report them to king Darius himself.

Based on accounting

In 1844 Luca Pacioli stated what we refer to as the first method in accounting, bookkeeping. He got this idea from the Venice merchants. Audit in the modern days acted as a guarantor to a financial statement, which was made using accounting principles. It means an auditor must understand accounting before they can become an auditor.


Audit 1.0 – 4.0

Audit 1.0 was auditing in the early industrial revolution, using the traditional way and no involvement of a computer. Using nothing but a pen and paper. It continues to be until the 1960s with IBM releasing computers for business purposes. Audit 2.0 started when UNIVAC was released in the mid-1950s but not commonly used before IBM released its business computer that comes with a reasonable price. Audit 3.0 was started with the use of a computer is a mass scale, from the database and the making of financial report


Audit 4.0

The advance in sensors, CPSs, IoT, and IoS  in the fourth industrial revolution, promotes a new intelligent, flexible, and secure factory: the ‘‘smart factory.’’ Hermann et al. (2015) imagined that smart factories would employ a completely new approach to production, in which smart products are identifiable and traceable with the capability of self-awareness and optimization, and the whole manufacturing systems are connected vertically with other business processes and horizontally with related parties outside of the factory. Shrouf et al. (2014) visualized the main components and processes in a typical smart factory (Figure 1). In a smart factory, production is initiated by receiving orders with customized requirements via a network that connects the entire factory and outside related parties. The smart factory then generates a manufacturing plan based on machines’ capabilities and status collected from the network, and autonomously guides raw materials and products throughout the production lines.

The smart factory produces smart products that integrate the functions of data processing, storage, and analysis. The smart products record and transfer their conditions and status, as well as customers’ behaviors and demands, to the factory to facilitate quality control and product design. The smart factory is also connected to suppliers to enable just-in-time inventory. Compared to the traditional manufacturing industry, the smart factories improve the flexibility of manufacturing, increase the automation of operations, enable mass customization and proactive maintenance, and connect all sectors in the value chain (Shrouf et al. 2014). Therefore, smart factories are becoming the core of Industry 4.0.

The dual worlds have their connections providing the key functionalities that the world of IoT and other technologies allow. Figure 4 shows how Audit 4.0 can provide a comprehensive assurance to a company by gathering and analyzing accounting and other business data from the entire business. The mirror world will continuously capture data that reflect the current status and performance of the organization. In addition, business processes will be monitored against predetermined rules to detect violations of key controls, and cross-verified via certain continuity equations (Kogan et al. 2014). Employees’ activities will also be captured by cameras and sensors to identify irregular or abnormal behaviors. Those data will all be linked to an organization’s ERP system to enable real-time accounting.

The mirror world will integrate real-time data from the entire organization into a data repository in the cloud. Auditors and other experts can create analytics models on top of the data repository in order to continuously detect anomalies, discover system faults, identify control inefficiency, and manage resources. Once an anomaly or fault occurs, the models will send an alert to auditors or management, who will promptly take action. Some control monitoring models can also be implemented in the individual equipment or facilities instead of the cloud if they only monitor local data generated or flowing through the machines. Such decentralization may dramatically reduce the workload in the cloud and improve overall efficiency.

Source :

  1. Auditing: An International Approach by Wally J. Smieliauskas and Kathryn
  2. Bewley, The evolution of auditing: An analysis of the historical development by Lee Teck-Heang and Azham Md. Ali
  3. JOURNAL OF EMERGING TECHNOLOGIES IN ACCOUNTING. By American Accounting Association. Vol.13 no.1. Spring 2016
  4. Evolution of Auditing: From the Traditional Approach to the Future Audit. By AICPA. November 2012.