Users' Perceptions and Trust in AI in Direct-to-Consumer mHealth: Qualitative Interview Study (Preprint)

• DOI: 10.2196/preprints.64715
• Publication Date: 2024-07-30T20:22:41Z
• Authors: Katie Ryan; Justin Hogg; Max Kasun; Jane Paik Kim
• Keywords: mHealth; Preprint

<sec> <title>BACKGROUND</title> The increasing use of direct-to-consumer AI-enabled mHealth (AI-mHealth) applications presents an opportunity for more effective health management and monitoring expanded capabilities. However, AI’s early developmental stage has prompted ethical concerns related to privacy, informed consent, bias, among others. While some these have been explored in stakeholder research AI-mHealth, the limited literature suggests that broader landscape considerations hold significance users may remain underexplored. </sec> <title>OBJECTIVE</title> Our aim was document explore perspectives regarding AI-mHealth applications. <title>METHODS</title> We conducted semi-structured interviews with (N=21) employed a qualitative descriptive design describe their perspectives. <title>RESULTS</title> Through analysis, three major categories nine subcategories describing users’ were identified. Users described attitudes toward impact on data (i.e., influences awareness management, value mental versus physical health, inevitability sharing); trust expert recommendations, technology companies, AI explainability); preferences relating information sharing type is collected, future uses data, accessibility information). <title>CONCLUSIONS</title> This paper provides additional context number previously posited or identified literature, including trust, explainability, sharing, revealed not documented, i.e., differentiation between cases, willingness extend empathy non-explainable AI. To our best knowledge, this study first apply open-ended, approach end <title>CLINICALTRIAL</title> addressed supplemental ongoing about ethics medicine (NCATS R01-TR-003505). obtained human subjects approval from Institutional Review Board Stanford University June 21, 2022 (#58118).