Applications of AI-Enabled Deception Detection Using Video, Audio, and Physiological Data: A Systematic Review

Artificial intelligence-enabled deception detection is an emerging tool for identifying dishonest behavior in a wide range of applications, from security and forensics to politics and lower-risk everyday interactions, addressing the pressing need for enhanced trust and security in an increasingly digital and interconnected world. However, to date, approaches to achieve deception detection with AI have not been evaluated by application area, leaving a disconnect between approaches leveraged and their potential real-world use. Thus, this paper provides a systematic review of application areas for AI-enabled deception detection approaches following the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) methodology. Specifically, we discuss 93 articles in detail, (1) identifying common applications of automated deception detection in the literature, (2) enumerating deception detection approaches by application area, and (3) describing publicly available datasets per application area. We also identify open challenges, such as the lack of datasets that support cross-domain deception detection research. By focusing on the application areas of automated deception detection, this review helps to contextualize the surveyed literature and understand the specific challenges and requirements associated with different domains. Further, by examining various application areas, researchers can tailor their approaches and techniques to address the unique characteristics and constraints of each domain. This targeted approach increases the practical relevance and applicability of deception detection methods in real-world scenarios.