人工智能价值链的法律型塑  

Legal Structuring of the Artificial Intelligence Value Chain

作　　者：李润生 Li Runsheng

机构地区：[1]北京中医药大学人文学院法律系

出　　处：《东方法学》2025年第2期63-78,共16页Oriental Law

基　　金：国家社会科学基金青年项目“人机协作视野下医疗人工智能临床应用法律问题研究”(项目批准号:22CFX051)的阶段性研究成果。

摘　　要：人工智能价值链是对人工智能组织过程的科学描述,是人工智能立法的重要线索。相较于有形产品和传统软件,人工智能价值链是一个整体协作、双轮驱动、动态发展和不透明的组织过程,这决定了人工智能的法律规制应当遵循协作规制、全生命周期规制和集中明确规制的核心理念。人工智能价值链的制度建构包含三个主要步骤,即提炼法律身份、选定中心节点和分配法律义务,以通用模型提供者、应用系统提供者、应用系统部署者和受影响者四种法律身份搭建人工智能价值链的组织骨架。其中,通用模型提供者和应用系统提供者位处中心节点,应当担负关键职责,推动构建公平、稳定和可信的人工智能价值链,应用系统部署者和受影响者也应当担负与其功能定位相匹配的义务和职责。The artificial intelligence value chain provides a scientific framework for understanding the organizational processes of AI and serves as a crucial guide for AI legislation.Distinct from tangible goods and conventional software,the AI value chain is characterized by holistic collaboration,dual-drive mechanisms,dynamic evolution,and inherent opacity in its organizational processes.These characteristics necessitate a legal regulatory approach for AI grounded in the core principles of collaborative governance,lifecycle-based regulation,and centralized yet clearly defined regulatory frameworks.Constructing an effective institutional framework for the AI value chain involves three primary steps:identifying key legal entities,delineating central nodes,and allocating corresponding legal obligations.The organizational skeleton of the AI value chain is constructed with four legal identities-general model providers,application system providers,application system deployers,and affected parties.General model providers and application system providers,situated at central nodes,should bear principal responsibilities in fostering the construction of a fair,stable,and trustworthy AI value chain.Application system deployers and affected parties should also assume obligations and responsibilities commensurate with their functional roles within the chain.

关 键 词：人工智能价值链 人工智能立法 生成式人工智能 通用模型提供者 责任分配 法律型塑 

分 类 号：DF0[政治法律—法学理论]