Factors of Trust Building in Conversational AI Systems: A Literature Review

Trust is essential for successful human-machine interaction. It is particularly important for conversational artificial intelligence (AI) systems in the service sector of the online world. This paper focuses on trust-building factors in conversational AI systems and explores strategies to strengthen trust. First, an overview of trust, AI, conversational AI systems, and related literature is provided before discussing related literature on the concept of trust and factors influencing user trust in human-computer interactions. Through a structured literature review, a concept matrix of several trust factors from the existing literature is presented. The findings highlight trust-building factors such as controllability, adaptability, transparency, intelligence, intimacy, empathy, engagement, anthropomorphism, security, brand perception, organizational trust, risk perception, personality traits, and expertise. Each factor has its importance and limitations in building user trust. For example, transparency enables a better understanding of users, but complex AI systems cannot be fully transparent, which leads to mistrust. Best practices from different domains highlight context-specific approaches that are essential for building trust in conversational AI systems. In addition, best practices, such as keeping control over the decision-making process and careful handling of sensitive data, were offered. The study highlights the importance of user trust in functionality, reliability, and security for the successful development and deployment of this technologies. Understanding user concerns and overcoming these barriers will lead the way for further development and innovation in this area.