Article Preview
Top1. Introduction
As firms and banks progressively expand their operations, more complex these become and to sustain the business there is one thing that should be taken into consideration: the risks. Human beings are constantly facing situations that require decision-making processes, each with different implications and therefore with different levels of risk (Pereira et al., 2021). The association between risk and reward is part of human common sense, and for that reason the reflection of risks is a decisive factor in any choice process, from the simplest things to the more complex ones (Weeserik & Spruit, 2018). Despite the controversial nature of the following statement, it is important to consider that the best decision is not necessarily the one that minimizes the risk, but the one that gives a better result for a certain degree of risk that one is willing to take. According to Drucker (2014), there are a lot of different types of business risks such as the financial risk, strategic risk, compliance risk, between many others. This investigation will focus on the Operational Risk.
The awareness of this type of risk began throughout the 90’s and is now being heavily explored and enriched. Operational Risk is commonly known as the risk resulting from the inadequacy or failure of both external events or internal processes, that could result in gain or lost earnings (Weeserik & Spruit, 2018). Thus, it is a risk associated to all activity domains, that can be presented through many forms such as frauds, human error, IT failures or natural catastrophes (Drucker, 2014; Diehl, 2014).
It is our main objective to study the implementation of Artificial Intelligence in OpRisk Management. We will address the success factor and down-sides of this implementation and pinpoint what can be the main drivers of success or unsuccess of the application of Artificial Intelligence in Operational Risk Management. In this investigation, we aim to fill the gap around the literature that covers these two areas, aiming to relate the progress in Artificial Intelligence with its possible application in the control of OpRisk. The article contributes to existing literature (i) by identifying success factors of these technologies are the cost reductions and improved efficiency. The reduction of human errors, higher precision in the controls done, and the fact them keeps employees motivated by eliminating the tedious tasks. Also, it was clear that the areas of risk identification and risk calculation are the ones that can have a higher return from the implementation of AI systems in OpRisk. Also, the software solutions available in the market are a success factors, as the majority seems to consider it easy to adapt to their needs. (ii) This study addresses the lack of knowledge of AI technologies and lack of investment in training staff as identified by Costa et al. (2020) and Pereira et al. (2021). Also, the lack of human rationality when novelties occur and possible future ethical and social concerns.