生成式人工智能产品责任承担研究
Research on the Responsibility Undertaking of Generative Artificial Intelligence Products

相较于传统物质性产品，生成式人工智能的特殊性使其难以被现行的产品责任所涵盖，致使民事责任承担机制的失调。物质形态并非判断生成式人工智能产品属性的唯一标准，从目的论角度出发，在现有产品责任框架内探讨生成式人工智能的责任承担问题，是合理划分举证责任、弥补损害、规制生成式人工智能发展的可行路径。生成式人工智能在产品责任适用中应当准确认定“生产者”角色，对通识的网络服务提供者概念进行修正，并制定明确的质量标准，结合警示说明义务和风险补救义务综合评判。本文以产品责任为基点，推动科技发展与立法更新的协调发展。
Compared to traditional material products, the particularity of generative artificial intelligence makes it difficult to be covered by current product liability, resulting in an imbalance in the mechanism of civil liability. The material form is not the only criterion for judging the attributes of generative artificial intelligence products. From the perspective of teleology, exploring the issue of responsibility for generative artificial intelligence within the existing product liability framework is a feasible path to reasonably allocate the burden of proof, compensate for damages, and regulate the development of generative artificial intelligence. Generative artificial intelligence should accurately identify the role of “producer” in the application of product liability, revise the concept of general network service providers, and establish clear quality standards, combined with the obligation of warning instructions and risk remediation obligations for comprehensive evaluation. This article takes product liability as the basis, promotes the coordinated development of technology and legislative updates.