Socially (dis)connected consumer trust in AI technologies: a systematic review and research agenda

Abstract

Purpose Loneliness, as a mental health epidemic alongside the artificial intelligence (AI) improvements, has received attention in the literature. Motivated by the need to belong, lonely consumers experience psychological states that influence their trust in AI, resulting in inconsistent findings in this emerging research stream. The purpose of this study is to provide a synthesized and structured overview of the state-of-the-art research focused on social (dis)connection, trust and AI. Design/methodology/approach A systematic literature review of 31 empirical studies was conducted following the PRISMA protocol. The review synthesizes findings from diverse research designs to identify theoretical themes related to socially (dis)connected consumer trust in AI. Findings Seven thematic domains emerged from the analysis – spanning social connection and belonging, self-differentiation, humanization, privacy considerations, social interpretation and human–computer and computer-mediated interaction. These domains inform a framework explaining how social (dis)connection shapes consumer trust in AI and outlining key theoretical, methodological and practical avenues. The review further introduces the compensatory-defensive trust in AI framework that clarifies when social disconnection heightens versus diminishes trust in AI. Originality/value This study presents a novel synthesis of literature on socially (dis)connected consumer trust in AI, proposing a new conceptual framework to guide future research. It emphasizes the importance of developing a comprehensive theory of lonely human trust in AI, exploring the nomological networks of related constructs and adopting methodological diversity. These contributions help advance academic understanding and inform ethical AI design sensitive to socially disconnected consumers.