Scenario Analysis Reveals How Generative AI Will Impact The Mental Health of Medical Students (Preprint)

BACKGROUND Introduction: Generative artificial intelligence (AI) is rapidly transforming medical education, while medical students continue to face high levels of stress, anxiety, and burnout. This dual pressure stemming from technological disruption and psychological vulnerability raises urgent questions about the future of training physicians. OBJECTIVE Introduction: Generative artificial intelligence (AI) is rapidly transforming medical education, while medical students continue to face high levels of stress, anxiety, and burnout. This dual pressure stemming from technological disruption and psychological vulnerability raises urgent questions about the future of training physicians. METHODS Methods: We used scenario analysis, a foresight methodology, to explore possible futures at the intersection of generative AI and medical students’ mental health. The exploratory scanning identified weak signals and emerging trends, which were clustered using a macro-meso-micro framework and interpreted through a multi-level perspective of socio-technical change. Scenarios were analyzed along two key dimensions: the extent of generative AI integration into medical curricula and the availability of mental health support. RESULTS Results: Four distinct scenarios emerged: Analogue Happiness (high support, low AI integration), Gen AI Paradise (high support, high integration), Disconnected Struggles (low support, low integration), and Gen AI Takeover (low support, high integration). Each illustrates different risks and opportunities for students’ digital readiness and psychological well-being. Discussion: The findings suggest that technological innovation and mental health support must co-evolve in medical education. Prioritizing one without the other risks producing either digitally unprepared or emotionally fragile physicians. Faculty readiness, ethical frameworks, and participatory curriculum design are critical to ensuring balanced integration. We formulated practical recommendations tailored to students, educators, and other stakeholders to guide balanced adaptation. CONCLUSIONS Conclusions: Generative AI is not an optional add-on but a transformative force, while mental health support is a prerequisite for competence. Institutions must rise to the dual responsibility of preparing students to become both digitally fluent and emotionally resilient. Failing to integrate these pillars constitutes not only a missed opportunity but an ethical failure with consequences for students, educators, healthcare professionals, and ultimately, patients.