A governança da inteligência artificial em sistemas de saúde públicos : um modelo proposto para a Diretoria de Saúde da Marinha.

Abstract

This study was motivated by the transformative potential of Artificial Intelligence (AI) in the healthcare sector and its increasing governance complexity. A specific segment of this broad topic was researched with the ultimate goal of proposing reflections and guidelines for the Brazilian Navy Health System (SSM) to promote safe and responsible innovation. The realities investigated were: the United Kingdom, the European Union, China, the United States of America, and Brazil. The investigative focus was on data privacy, explainability, biases, and human oversight. Singularities were highlighted, reflecting distinct structures in the pursuit of a balance between innovation and regulatory rigor. In addition to this investigation, the study aimed to diagnose the reality of the SSM and its digital maturity, which revealed significant potential, but mitigated by challenges such as data fragmentation and lack of interoperability, gaps in specialized human resources, and the absence of an internal regulatory framework, all within an organizational culture that prioritizes sustainment over innovation. The combination of investigation and diagnosis led to the following proposals: alignment of the strategic vision with modernization; adoption of risk-based governance; strengthening of ethics, privacy, and individual rights; and management and quality for the benefit of military health big data. Furthermore, the following were emphasized: transparency and explainability of algorithms; proactive bias mitigation; the need for an AI development and implementation plan focused on capacity building; responsibility and human oversight; AI as support, not a substitute for professional judgment; and cybersecurity and data sovereignty. It was evident that AI can be a catalyst for overcoming systemic challenges, such as the scarcity of skilled labor and inefficiencies in complex healthcare systems. A clear and well-founded roadmap for AI implementation was presented, balancing innovation, safety, privacy, and ethics. The integration of AI into the SSM proves to be a complex, yet urgent and transformative undertaking, requiring a multifaceted, ethical approach with a focus on human capital, aiming for more efficient, safe, and humane healthcar