The Impact of Applying Artificial Intelligence on the Quality of Enterprises Decision-Making

Abstract

Artificial Intelligence (AI) is becoming more heavily utilised in enterprise decision-making process, and it is being speculated that AI is on course to becoming capable of fully supporting autonomous decision-making. However, there are significant, non-replicable, human elements in a decision-making process, which are inherently absent in AI systems. This raises the question of the extent to which AI is able to support decision-making in its current form. AI has been heavily researched in the literature, but its role in enterprise decision making have only been very recently realised. It is important to realise where AI assisted decision-making process stands today. Most of the research conducted on AI decision-support systems have focused on the possibilities, opportunities, and challenges of being integrated into an enterprise decision-making process. What is lacking in the literature is an aggregate of the findings so that future research directions and investment opportunities can be better explored. In order to fill this gap, this research qualitatively analyses the very recent existing literature on AI’s contribution to enterprise decision-making process. The methodology applied is secondary data analysis in the form of a literature review conducted using the PRISMA method. Seminal research papers have been reviewed, and their findings analysed and aggregated. The findings indicate that AI is most utilised in data processing as this requires little input from humans other than feeding large, fast-paced, real-time data. The findings further show that beyond the decision-making process, AI is used in market prediction, and in healthcare enterprises for making medical oriented decisions. Finally, the most common AI algorithms that are utilised to help in enterprise decision-making process are artificial neural networks and swarm intelligence. Based on the findings, this research also suggests future research implications for academics.