Trust in Medical AI: The Case of mHealth Diabetes Apps.

Over the past few years, mobile health applications (mHealth apps) have gained popularity regarding the management of chronic conditions, such as diabetes, as they are considered to enhance follow-up and treatment. Indeed, these applications are powerful tools that support individualised pharmaceutical and non-pharmaceutical care by remotely monitoring the patient's health status in real-time. Nevertheless, concerns have been raised regarding trust and trustworthiness towards their use, in particular, when they are powered by Artificial Intelligence (AI). Trust in AI is a multifaceted notion that encompasses aspects such as reliance, risk, as well as ethical principles (e.g., respect for human autonomy and prevention of harm). In this paper, we provide an analysis of trust and trustworthiness in AI-powered mHealth apps for diabetes through the lens of philosophy and risk analysis to promote the patient's well-being.