智能医疗算法审查的融合型立法研究

Research on Convergent Legislation for Intelligent Medical Algorithm Review

算法不应成为法外之域,算法审查亦不应采取机械化、概括式或单向度规制的路径,而应立足“领域法”的特殊性,通过嵌入式的智能医疗算法审查融合立法,精准识别智能医疗算法系统的潜在安全隐患,通过传统风险与新兴风险立法的融合,防控智能医疗算法决策引发的叠加型医源型与药源型安全风险。通过管理型立法与促进型立法的融合,处理智能医疗算法分发对健康公平产生的双向影响效应;通过技术治理与组织治理领域立法融合,有效应对医疗健康及生物识别信息安全隐患;通过控制者义务与外部算法审查相结合,增设智能医疗算法决策风险最小化的命令性规范、智能医疗算法应用的动态分级授权制度、预警及诊疗算法模块信息爬取的限制性规范,为人民生命健康提供有力法治保障。

Algorithms should not become the domain outside the law, and algorithm review should not adopt the path of mechanization, generalization or one-way regulation. Instead, based on the particularity of "domain law", it should accurately identify potential security risks of intelligent medical algorithm system through embedded intelligent medical algorithm review and fusion legislation. Through the integration of traditional risk and emerging risk legislation, we can prevent and control the superimposed medical source and drug source security risks caused by intelligent medical algorithm decisions. Through the integration of regulatory legislation and promotional legislation, we can deal with the two-way impact of intelligent medical algorithm distribution on health equity. Through the legislative integration in the field of technical governance and organizational governance, we can effectively address the security risks of medical health and biometric information. Through the combination of controller’s obligations and external algorithm review, adding mandatory specifications, we can minimize the risk of intelligent medical algorithm decision-making, provide a dynamic hierarchical authorization system for the application of intelligent medical algorithm, and restrict information crawling of early warning and diagnosis and treatment algorithm modules to provide a strong legal guarantee for people’s life and health.