Toward Autogenic mHealth: A Shared Vision for Self-Directed Health Support
On November 9, 2026, the mDOT Center convened a workshop on Autogenic mHealth: Self-directed Monitoring, Management, and Intervention to explore a fundamental shift in digital health: moving from systems that primarily deliver expert-defined support to systems that help people understand, direct, and act on their own health in partnership with clinicians, caregivers, communities, and researchers.
The workshop brought together perspectives from behavioral science, artificial intelligence, mobile health, wearable sensing, clinical care, human-centered design, and systems science. Across the day, participants considered what becomes possible when generative AI, health foundation models, wearable data, and personalized digital twins are designed not simply to predict or prompt, but to support people’s agency, learning, and self-directed care.
A defining theme of the day was human agency. The promise of autogenic mHealth is not to replace people or care teams with technology; it is to develop tools that help people become more informed and capable partners in their own health journeys. This includes respecting personal goals, making system interpretations transparent and correctable, and recognizing the social and community conditions that shape health.
From Adaptive Interventions to Autogenic mHealth
Participants began by distinguishing established just-in-time adaptive interventions from the emerging idea of autogenic mHealth. Traditional adaptive interventions use available data to select from a predefined set of options at an appropriate moment. Autogenic systems point toward a more open-ended future: one in which people can ask questions, express goals and lived experience in their own words, correct a system’s understanding, explore possible actions, and remain active participants in shaping their care.
This vision does not eliminate the role of clinicians or scientific evidence. Instead, it reframes digital health as a connected process involving individuals, caregivers, care teams, intervention components, system models, and adaptive tools. Participants emphasized that the most valuable systems will help these elements communicate and work together while preserving meaningful human oversight.
Building Technology Around Agency, Context, and Community
A central theme of the workshop was that health technologies should be responsive to what people value—not only to what a model can infer. Participants discussed systems that allow individuals to set their own goals, decide what information to share, challenge incorrect interpretations, and learn through optional personal experiments.
The conversation also made clear that many health challenges cannot be addressed through individualized nudges alone. Community conditions, access barriers, caregiving responsibilities, and structural inequities shape people’s opportunities to act on health information. Autogenic mHealth must therefore be designed with privacy, community stewardship, and collective action in mind, alongside personalization.
Examples discussed during the workshop illustrated both the promise and the practical realities of this approach. Participants considered personal tracking and self-experimentation, context-aware support for diabetic foot care, and tools that could help caregivers understand and respond to nighttime disruptions. These examples highlighted the potential for technology to help people recognize patterns and test strategies that matter to them, while also underscoring challenges related to affordability, connectivity, technology burden, and unequal access.
The Technology Under the Hood
The afternoon sessions examined the computational foundations that could help enable autogenic mHealth. Discussions included self-supervised wearable models, time-series foundation models, multimodal AI, and personalized digital twins.
Foundation models may help make sense of complex, incomplete, and noisy streams of wearable and real-world data. Digital twins may offer a way to combine individual measurements with broader scientific knowledge, supporting exploration of health trajectories and possible interventions. Yet participants were clear that these technologies remain works in progress. Data scarcity, missingness, device differences, opaque internal representations, and uncertainty about whether population evidence transfers to an individual all remain important research challenges.
The group organized these opportunities into three practical categories: technologies that may be ready to use now, areas requiring joint development, and fundamental research bottlenecks that should shape future work.
Safety, Evidence, and Clinical Alignment
The workshop placed safety and evidence at the center of the conversation. Participants explored the role conversational agents could play in gathering information, supporting engagement, and helping people reflect on their experiences. At the same time, they cautioned against treating AI-generated guidance as a substitute for clinical judgment. When health recommendations are involved, clinicians must retain responsibility for care decisions.
Participants also considered how to rigorously evaluate systems that change over time and respond to individual choice. Existing approaches such as ecological momentary assessment remain valuable for structured scientific evaluation, even as new conversational approaches may offer richer forms of interaction. The challenge ahead is to distinguish intervention design from evaluation design, measure meaningful outcomes and mechanisms, and ensure that system inferences remain correctable rather than treated as fact.
Safety, liability, privacy, trust, ethics, bias, social justice, and regulation emerged as closely linked priorities. These questions will be essential to determining not only whether autogenic mHealth systems work, but whether they deserve to be trusted.
A Collaborative Path Forward
The workshop concluded with momentum around a shared vision paper on autogenic mHealth. The proposed paper will define the concept, distinguish it from existing intervention frameworks, describe motivating use cases, and identify research and technology priorities. Participants also began organizing collaborative work around an overall framework, digital twins, study design and evaluation, and safety and governance.
As the mDOT Center moves into its next phase, the ideas advanced through this workshop will help guide future research on technology that is not only intelligent and personalized, but also supportive of the people and communities it is intended to serve.
Workshop Objectives:
Autogenic mHealth is an emerging paradigm for personalized digital health in which individuals work continuously with intelligent systems to monitor, understand, and manage their own health. Rather than relying solely on expert-designed and delivered interventions, Autogenic mHealth shifts toward participant-driven health management, with AI-enabled systems interpreting personal physiological, behavioral, and contextual data while incorporating scientific and clinical knowledge to support individualized decisions.
Advances in biosignals, multimodal sensing, foundation models, generative AI, digital twins, and agentic AI are creating the technological foundation for this shift. These capabilities can enable systems to move beyond passive monitoring and static recommendations toward continuous partnerships that help individuals discover what to monitor, understand patterns in their health and behavior, explore potential strategies, and learn from the results over time.
At its core, Autogenic mHealth represents a shift from expert-driven intervention to participant-driven self-management:
Monitor → Understand → Explore → Act → Learn → Adapt
The goal is not to replace researchers or clinicians, but to extend scientific and clinical knowledge into an ongoing partnership with individuals, enabling more informed, personalized, and adaptive health decisions in everyday life.
Traditional mHealth has largely centered on interventions designed by experts and delivered to individuals. Researchers and clinicians typically identify what should be measured, determine the desired outcome, develop an intervention, and establish when and how it should be delivered. Even personalized approaches generally operate within a framework defined in advance.
Autogenic mHealth shifts this model toward participant-driven health management. Instead of simply receiving predetermined recommendations or interventions, individuals work with intelligent systems to understand their own health, identify meaningful patterns, explore potential strategies, and learn from their experiences. The system continuously integrates personal data with scientific and clinical knowledge, allowing the process to evolve as the individual’s goals, context, and health change.
This represents a shift from:
Expert-driven → Participant-driven
Predetermined interventions → Self-directed strategies
Static recommendations → Continuous learning
Passive monitoring → Active understanding
Intervention delivery → Ongoing health partnership
The distinction is fundamental: Autogenic mHealth is not simply about making existing interventions more intelligent. It is about enabling individuals to become active participants in the process of discovering, understanding, and managing their own health.
The workshop will explore the emerging frontier of Autogenic mHealth and its potential to transform personalized digital health. Participants will examine how biosignals, multimodal sensing, foundation models, generative AI, digital twins, and agentic AI can work together to create intelligent systems that continuously learn from an individual’s physiological, behavioral, and contextual data. Particular attention will be given to how these technologies can move beyond passive monitoring and static recommendations toward systems that actively support individuals in understanding and managing their own health.
The workshop will identify opportunities and challenges associated with moving from expert-driven interventions to participant-driven self-management. Participants will explore how intelligent systems might help individuals discover what to monitor, understand relationships between behavior and physiology, evaluate potential strategies, and adapt their approach over time. Discussions will also examine the scientific and practical challenges involved in translating this vision into effective and usable systems, including the interpretation of complex multimodal data, personalization, continuous learning, and integration of scientific and clinical knowledge.
The workshop will address the ethical, methodological, and technical challenges associated with enabling individuals to use intelligent systems to make decisions about their own health. Discussions will consider issues of privacy, autonomy, consent, trust, transparency, safety, and responsible use of AI in systems that continuously interpret personal health data and provide individualized guidance. Participants will also examine technical challenges associated with reliable sensing, multimodal data integration, model reasoning, personalization, adaptation, and the safe deployment of increasingly autonomous AI systems in real-world health settings.
A pivotal goal of the workshop is to develop a strategic research roadmap for Autogenic mHealth. Through structured discussions and working sessions, participants will identify foundational research questions, theoretical gaps, methodological needs, technological priorities, and opportunities for translation. The workshop will seek to define the key scientific challenges that must be addressed to establish Autogenic mHealth as a rigorous and responsible area of research and to identify priorities that can guide future studies, technologies, and collaborations over the coming years.
The closing session of the workshop will provide an opportunity to synthesize the ideas, opportunities, challenges, and priorities identified throughout the day. Participants will discuss concrete next steps for advancing the emerging Autogenic mHealth research agenda, including opportunities for collaboration, new research directions, and development of shared resources and frameworks. The workshop will seek to build an interdisciplinary community of researchers committed to establishing the scientific foundations of participant-driven, AI-enabled health management and advancing the field beyond the workshop.