I Used AI for My Chronic Illness for a Year. Here's What Went Wrong. (Tjasa Zajc, Agentic Patient)

In this episode of 'Faces of Digital Health,' host Tjasa Zajc shares her personal journey of using AI to manage her chronic illness over the past year, revealing both the transformative potential and profound risks of patient-driven AI use. Initially captivated by AI's ability to provide answers and comfort during a health crisis, she experienced a rollercoaster of trust and confusion as different AI models gave conflicting advice—some urging rest, others interpreting her activity as a sign of recovery. This led her to develop the concept of the 'agentic patient': a proactive, critically thinking individual who uses AI not to replace clinicians, but to become a more informed and effective partner in their care. She identifies three patient archetypes—minimizers, cyberchondriacs, and informed collaborators—and stresses the urgent need for digital literacy, community-driven education, and systemic support to prevent harm. The conversation highlights how patients are already using AI independently, often without clinician awareness, creating risks when information is hidden or misinterpreted. Zajc argues that healthcare systems must evolve to respect patient agency, integrate patient-generated data, and promote shared decision-making through accessible, engaging education—like Finland’s open 'Elements of AI' course. Ultimately, she envisions a future where patients learn from each other in webinars and workshops, co-creating best practices to navigate AI responsibly and collectively transform healthcare from the ground up.