Artificial intelligence chatbots are now a familiar presence in online retail, answering questions, recommending products and handling routine service tasks at a scale no human team could match. Yet a new systematic review of the research literature suggests that the technology’s future in shopping depends less on what chatbots can do than on whether customers believe their data is safe. The review, published in Frontiers in Artificial Intelligence, synthesises 54 studies published between 2016 and 2026 to map how privacy, security and trust interact to shape the acceptance of AI-powered self-service chatbots in retail environments.
The study was conducted by Steven Kayambazinthu Msosa of Mangosuthu University of Technology in South Africa and followed the PRISMA framework, the standard protocol for systematic reviews in health and social science research. Searches of Scopus, Web of Science, ScienceDirect and IEEE Xplore yielded a final set of 54 scholarly works that met the review’s inclusion criteria. Rather than measuring customer attitudes directly, the review aggregates findings from a decade of empirical and conceptual studies, aiming to integrate what the author describes as fragmented insights into a single explanatory framework for retail chatbot adoption.
The central finding is that privacy remains the leading source of user resistance. Across the reviewed literature, shoppers consistently worry about how chatbots collect personal information, whether the systems are transparent about what data they gather, whether consent is obtained in a meaningful way, and what happens to information once it is shared with third parties. The review indicates that these concerns about transparency, consent and third-party data sharing are a primary reason shoppers hesitate to engage with automated service channels.
Security forms the second pillar of the analysis. The reviewed studies identify vulnerabilities that include the risk of data breaches and insufficient encryption mechanisms protecting the information exchanged between customers and chatbot systems. These technical weaknesses, the review finds, further undermine confidence in chatbot platforms.
Trust emerges as the variable that binds the framework together. According to the review, trust is an important influence on both behavioural intention, meaning a shopper’s stated willingness to use a chatbot, and user satisfaction once the interaction has taken place.
From these findings the author derives a set of implications for organisations deploying chatbots in retail. The review highlights the need for transparent data practices, so that customers understand what is collected and why; stronger cybersecurity infrastructure, to reduce breach risk and protect data in transit; and ethical AI design, so that automated systems treat user information in ways customers can verify. The argument is that sustainable user trust depends on these organisational commitments rather than on conversational polish alone.
The review also identifies gaps that current research has not filled.
As with any systematic literature review, the conclusions describe patterns in published scholarship rather than new measurements of shopper behaviour. The 54 included studies were selected through the PRISMA process from four academic databases, and the strength of each finding depends on the quality and design of the underlying research.
Subject of Research: A systematic literature review of privacy, security and trust in AI-powered self-service chatbots in the retail sector
Article Title: Unravelling the dynamics of privacy, security, and trust in AI-powered self-service chatbots in the retail sector: a systematic literature review
Article References: Msosa, S. K. (2026). Unravelling the dynamics of privacy, security, and trust in AI-powered self-service chatbots in the retail sector: a systematic literature review. Frontiers in Artificial Intelligence, 9, Article 1958107. https://doi.org/10.3389/frai.2026.1958107
Image Credits: AI Generated
DOI: 10.3389/frai.2026.1958107
Keywords: artificial intelligence, chatbots, retail, privacy, security, trust, systematic review, PRISMA, data transparency, customer adoption, consumer behaviour, ethical AI
News Source: Denise Maddox. (November 4, 2026). Privacy, Security and Trust Shape Retail Chatbot Adoption, Review Finds. Scienmag.
