人工智能产品致害风险及其侵权责任规制

Regulations on the Risk of Harm Caused by Artificial Intelligence Products and Its Tort Liability

人工智能产品具有自主决策能力和深度学习能力,在蕴含巨大发展潜力的同时也存有不可避免的风险,亟须从法律制度层面作出有效规制。在注重损害预防的理念之下,要求制造商在产品研发阶段预先设计出符合法律和伦理的算法程序规则,并将之嵌入人工智能系统以尽可能降低致害风险。以智能机器人和智能网联汽车等为代表的人工智能产品致害事故主要归因于其缺陷问题,可以统一纳入产品责任制度的范畴。人工智能产品致害侵权责任应当坚守严格责任原则的基本立场,明晰智能软件的产品属性并依据违反既定设计标准和故障理论判定其缺陷问题,扩充责任主体为生产者、销售者、研发者以及改造者,对于发展风险抗辩仍需保留并注意弥补可能带来的损害。为了应对大规模人工智能产品侵权责任纠纷和及时救济受害人,还应当充分发挥责任保险机制的配套作用,构建多层级的保险结构,探索个性化的保险模式。

Artificial intelligence products have independent decision-making ability and deep learning ability,which contain huge development potential but also have inevitable risks,which need to be effectively regulated from the level of legal system.Under the concept of focusing on damage prevention,manufacturers are required to design legal and ethical algorithmic programming rules in advance during product development,and embed them into artificial intelligence systems to minimize the risk of harm.The typical artificial intelligence products such as intelligent robots and intelligent connected cars are mainly caused by their defects,which can be uniformly included in the scope of product liability system.The liability for infringements caused by artificial intelligence products should adhere to the basic position of the principle of strict liability,clarify the product attributes of intelligent software,and determine its defects based on violations of established design standards and malfunction doctrine,and expand the main body of responsibility to manufacturers,sellers,developers and reformers.In addition,the defense against development risks still needs to be retained and attention should be paid to make up for possible damage.In order to deal with large-scale artificial intelligence product tort liability disputes and provide timely relief to victims,it is also necessary to give full play to the supporting role of the liability insurance mechanism,build a multi-level insurance structure,and explore personalized insurance models.