Abstract

The use of artificial intelligence (AI) to promote healthy behaviors has increased in recent years. This study’s purpose was to explore how AI devices have been used to promote physical activity and how functional components of these devices align or could be improved based on past successes of theory-based interventions to increase physical activity behavior. This systematic scoping review was conducted following the PRISMA extension for scoping reviews; databases searched included PsycINFO, Web of Science, Scopus, and CINAHL. Included articles were published between 2018 and 2023, peer-reviewed, conducted among human subjects, and focused on understanding the use and functional components of AI to promote physical activity. Of the 2,526 total articles exported, fifty-two were included in the study. Thematic and content analysis employing three cycles of coding was performed. The following AI devices promoted physical activity: health gamification systems (35%); mobile applications (29%); work, education, and home-based exercise programs (15%); virtual assistants (15%); virtual reality (13%); wearable technology (13%); robots (12%); and other devices (6%). Devices that incorporate rewards and incentives to encourage exercise increased user motivation to exercise, and interventions tailored toward specific populations such as youth and young adults or older adults can increase exercising behavior. Aligned with the constructs of social and behavior change theories such as the Social Cognitive Theory, findings supported the premise that individuals who were motivated to participate in physical activities, had pro-exercising attitudes, were exposed to an environment that facilitated the behavior, and were able to achieve optimal self-efficacy successfully increased exercising behavior.