人形机器人自主侵权的责任认定  

Determination of Liability for Autonomous Tort of Humanoid Robots

作　　者：谢琳 Xie Lin

机构地区：[1]中山大学法学院

出　　处：《东方法学》2024年第3期77-87,共11页Oriental Law

基　　金：国家社科基金项目“私法适用中再分配问题的法律经济学研究”(项目批准号:22BFX177)的阶段性研究成果。

摘　　要：因人工智能系统的不可解释性、不可预测性和自适应性,人形机器人自主侵权责任面临难以认定的困境。人形机器人自主致人损害属高风险人工智能领域,对此人形机器人提供者应承担产品责任,履行产品安全可控保障义务;大模型提供者仅需履行充分的透明度告知义务,与人形机器人提供者承担共同侵权责任;人形机器人使用者须按照说明正确使用产品,其承担的侵权责任不宜过度偏离原有类型。通过层次化确立信息披露、可反驳的过错和因果关系推定,不再要求过错与因果关系具备一一对应性,可使受害者获得公平救济的同时兼顾产业发展。Due to the inexplicability,unpredictability,and adaptability of artificial intelligence systems,the determination of liability for autonomous tort of humanoid robots faces challenges.Autonomous harm caused by humanoid robots falls within the high-risk domain of artificial intelligence.Providers of humanoid robots should bear product liability and fulfill obligations to ensure product safety and controllability.Providers of large models only need to fulfill sufficient disclosure obligations,sharing joint tort liability with providers of humanoid robots.Users of humanoid robots must use the product correctly according to the instructions,and their liability for tort should not deviate excessively from the original type.By hierarchically establishing information disclosure,rebuttable fault,and causal relationship presumptions,without requiring a one-to-one correspondence between fault and causation,fair relief for victims can be achieved while also considering industrial development.

关 键 词：人形机器人 自主侵权 侵权责任 大模型 过错 因果关系推定 

分 类 号：DF522[政治法律—民商法学]