The Role of AI Agents in Spreading and Detecting Fake Online Reviews: A Systematic Review: An Abstract

Abstract

Researchers have underlined the need for a better understanding of the variables that contribute to and undermine the process of fake reviews detection, to help business, government, and non-profit organizations design more effective strategies (Grazioli & Jarvenpaa, 2000; Xiao & Benbasat, 2011). The purpose of this paper is to offer a comprehensive literature review on the role that AI-based technology plays in the creation, diffusion, as well as detection of fake online consumer reviews, and to propose a theoretical framework for future research. There is a developing literature on digital deception detection based on artificially intelligent tools, including machine learning and automated deception classifiers in analyzing textual and contextual indicators of manipulation (Anderson & Simester, 2014; Munzel, 2015). In this context, artificial intelligence represents a resource to automate business processes in online targeting, engaging consumers, as well as gain insights from data, analyzing numbers, text, voice, faces, and images (Davenport & Ronanki, 2018; Grewal et al., 2020; Shankar, 2018; Wirth, 2018; Yadav & Pavlou, 2020). AI agents have the potential to be successfully employed by both consumers and marketers in the process of spreading and detecting deceptive reviews, through technologies like social bots and machine learning algorithms that can make the process of manipulation more effective but can also help in its detection. Drawing from the theoretical framework based on the Persuasion Knowledge Model (Campbell & Kirmani, 2000; DeCarlo, 2005; Friestad & Wright, 1994) and the sinister attribution error in consumer judgments (Kramer, 1998; Main et al., 2007), we are proposing a framework for future research that focuses on both the intervention of AI-agents in the fake review process, as well as the changing equilibrium in consumer and business outcomes as a result of fake review use. This framework synthesizes the process of diffusing fake online reviews and their impact on consumers and organizations. This paper represents the first attempt to provide an extensive and critical review on the topic of AI-based manipulation and detection. A systematic review of the literature helps us advance our current knowledge of online reviews in several impactful ways. (1) It helps us identify AI-agents manipulation and detection mechanisms applied to understand deceptive online reviews, (2) It allows us to shed light on the unique characteristics of online reviews that allow for its manipulation and detection, and (3) it gives the ability to study the inter-relationship of literature themes using PKM to propose a direction for future research.