JMIR Publications

ABSTRACT

Background:

Artificial Intelligence (AI) is transforming health promotion by enabling novel data-driven approaches to persistent public health challenges. Yet, the field lacks a structured framework to organize how AI applications operate across different contexts and stakeholders.

Objective:

To structure this evolving landscape, we propose the AI–Health Promotion Framework, which classifies applications along two dimensions: physicality (tangible vs. intangible) and stakeholder orientation (patient- vs. provider-centered). This lens captures how AI, whether embodied in devices such as wearables or expressed through algorithms and applications, interacts with different health stakeholders.

Methods:

We conducted a computational literature review using Natural Language Processing and unsupervised machine learning. A total of 6,328 scientific publications were analyzed. Topic modeling was performed to identify thematic structures, and scientometric analysis was applied to map research clusters and assess intellectual linkages.

Results:

We identified eight topics from the literature, which clustered into two broad domains of AI in health promotion. The first domain, ethical and societal considerations, includes research on data governance and privacy, algorithmic bias and fairness, policy and regulation, and public trust in AI. The second domain, AI-enabled health interventions, covers studies on mobile health applications, conversational agents and chatbots, wearable and IoT devices, and virtual or digital interventions.

Conclusions:

The AI–Health Promotion Framework advances theoretical understanding by extending AI experience models to include physicality and stakeholder orientation. It also provides practical insights for digital health policy, technology development, and public service design, offering a structured roadmap for the equitable and effective integration of AI in health promotion.