Abstract: Recent advancements in artificial intelligence (AI) within healthcare and public health sectors have spurred the development of AI4Health/Medicine technologies, introducing innovative dynamics into the healthcare system. While underscoring the efficiency, advanced capabilities, precision, and personalization of medical services enabled by AI, acknowledging the centrality of ethical issues to the sustainable development of AI4Health/Medicine is crucial. Without proper regulation, the ethical risks associated with AI4Health/Medicine could engender severe negative consequences upon its widespread application in clinical and social contexts. Ethical risks in technology, leading to damage to social trust and harm to stakeholders such as patients, physicians, and smart health product consumers, are intolerable without stringent ethical regulations. Therefore, it is essential to establish an effective governance system to ensure the ethical integrity and social viability of AI4Health/Medicine technology. Integrating technological ethics and addressing social concerns within the AI4Health/Medicine framework, assessing potential risks from technological innovation, and enhancing governance over ethics are imperative. This article builds upon existing academic findings concerning ethical issues in AI4Health/Medicine, with a focus on governance over ethics as the core subject. By examining current studies on the governance over ethics of AI4Health/Medicine, this paper analyzes its challenges and limitations. Building on this analysis, the paper explores the difficulties and dilemmas of governance over ethics in AI4Health/Medicine, proposing a theoretical framework for achieving development goals. This study reveals that the innovation and application of AI4Health/Medicine conceal unforeseen and manageable ethical risks and issues. Furthermore, the disruptive changes in healthcare and public health due to AI4Health/Medicine technology render ethical issues and governance indispensable to the ongoing transformation of medical intelligence. The governance over ethics of AI4Health/Medicine differs from that of other AI forms. The ethical issues of AI4Health/Medicine exhibit three distinct characteristics: contextual sensitivity, non-closure, and unique trust relationships. Consequently, the management of AI4Health/Medicine's ethical issues requires systematic, platform-based governance projects for comprehensive support. Given the social concerns regarding AI4Health/Medicine, collaborating with effective governance mechanisms to prevent the pitfalls of medical technologism, avoid technology's alienation due to ethical issues, and uphold a people-oriented approach is crucial. This suggests that developing a system for the governance over ethics of AI4Health/Medicine aims to foster sustainable innovation and ensure the transformation of medical intelligence aligns with humanistic care principles. Building on this, the article outlines four practical approaches centered on the ethical values of AI4Health/Medicine: (1) The ethical value of technology constitutes the cornerstone of AI4Health/Medicine's governance over ethics, transforming ethical regulations into value objectives pursued consistently throughout the governance process; (2) Collaboration among governance entities is centered on shared ethical values, eschewing disputes over the maximization of interests; (3) Governance over ethics in AI4Health/Medicine requires that value requirements transcend mere conceptualization to become institutionalized, guaranteeing their applicability and stability; (4) The achievement of governance over ethics in AI4Health/Medicine necessitates continual adjustments in light of evolving scenarios and technological progress. This allows us to guide, adapt, and realize a "goodness-oriented" ethical vision, harnessing governance over ethics initiatives during the transformation of medical intelligence, thereby advancing the objective of enhancing AI4Health/Medicine technology.

摘要： 医疗健康是人工智能极具应用与市场潜力的分殊领域，但也带来不少棘手的伦理风险亟待治理应对。对医疗人工智能伦理实施专项治理，其因是医疗人工智能虽与一般性人工智能伦理问题“同源同根”，但特殊的应用语境、非封闭性条件以及医疗语境下的信任需求，又使医疗人工智能的伦理问题更为独特、切身和敏感。目前对医疗人工智能伦理治理的认识并不充分，缺乏可靠的治理依据、治理对象模糊、机制尚不健全以及治理意识的匮乏，加之技术场景变化难以确立、多元伦理诉求难以调和、治理责任分配不清和治理决策缺少实证依据，进一步增加了医疗人工智能伦理的治理难度。为此应扬弃既往“问题跟随”的堵漏式治理，尝试发展一种体系化的伦理治理模式，以医疗人工智能伦理价值为核心，构筑动态、稳健、可持续的医疗人工智能伦理风险的系统性治理机制，从而实现对医疗人工智能伦理问题的有效治理。