Confiança do usuário em chatbots: uma revisão bibliográfica

Abstract

Artificial intelligence has been undergoing a rapid process of evolution, being present in the daily lives of people in the most varied areas. It refers to the automation of activities that are commonly associated with human cognition, such as decision making, learning, and problem solving. This technology has been growing in popularity especially in the field of chatbots, applications that use natural language processing to simulate a conversation with humans. However, its adoption was lower than initially expected, probably due to a lack of focus on user needs. User trust in this technology is essential not only for its adoption, but also for encouraging cooperation and continued use, especially in contexts where there is a high perception of risk. The rapid growth of this technology in Brazil has not yet been accompanied by knowledge of the factors that influence trust in it. Thus, this study proposed to explore these factors through a literature review, identifying, describing and categorizing relevant studies. Nine studies were selected for review, from 2001 to 2021. The identified factors were separated into three major categories: (1) chatbot-related factors, (2) user-related factors, and (3) environmental or contextual factors. Chatbot-related factors were divided into two other categories: reliability, which includes factors related to chatbot performance, such as expertise, responsiveness, credibility, and ease of use; and anthropomorphism, which encompasses factors referring to the human likeness of the chatbot, such as appearance, social presence, human warmth and communication strategies. User-related factors include previous usage experience, the user's propensity to trust the technology, their demographics and sociocultural background, as well as different personality dispositions. Finally, environmental factors comprise the context of the task and service provided, the perceived risk, as well as security and privacy concerns. By bringing an overview of the factors that influence user trust in chatbots, the study contributes to the deepening of this knowledge in brazilian academic production and also helps designers to better understand what users' needs are when interacting with this technology.