The Agentic Patient 4: Finding Insurance and Red Team Analysis

In this episode of *Faces of Digital Health*, host Tiasho explores the transformative role of AI in patient empowerment through the story of Dmitri Giannikopoulos, Chief Innovation Officer at Red AI and a long-term patient living with multiple sclerosis. Dmitri shares his personal journey of navigating a fragmented U.S. healthcare system, from a decade-long diagnostic delay to managing complex insurance decisions. He reveals how he used AI—specifically large language models like ChatGPT and Anthropic—to conduct a red team analysis of insurance plans, ultimately identifying a high-deductible plan that saved him thousands in out-of-pocket costs despite higher upfront payments. His approach involved feeding massive amounts of clinical and insurance documentation into AI, using adversarial prompting to challenge assumptions and uncover hidden financial traps. The episode underscores the critical importance of detailed, context-rich prompts, skepticism, and verification when using AI for health decisions. Dmitri also reflects on the broader implications of AI in healthcare, advocating for responsible deployment—especially in denial workflows—where human oversight is essential. He draws parallels between AI and nuclear power: immensely powerful but requiring careful governance. The discussion concludes with practical advice for patients: avoid vague questions, red team AI outputs, and engage with patient and family care councils to shape more humane, patient-centered systems. The episode is a compelling case study in how informed, agentic patients can leverage AI to navigate systemic complexity. Key takeaways include: 1) Never ask simple AI questions—always provide full context and documentation; 2) Use adversarial prompting (e.g., 'red team this') to challenge AI’s assumptions; 3) Verify AI outputs against your own records and experience; 4) Engage with patient councils to influence system design; 5) Recognize that AI is a tool for augmentation, not replacement, especially in high-stakes medical decisions. The overall tone is cautiously optimistic, emphasizing empowerment through knowledge and critical thinking rather than blind trust in technology.