ChatGPT生成式人工智能侵权责任研究

Research on Infringement Liability of ChatGPT Generative Artificial Intelligence

以ChatGPT为代表的生成式人工智能的侵权责任,应从人工智能自身的法律定位问题以及相关主体侵权责任认定上进行探究。目前,以ChatGPT为代表的生成式人工智能侵权责任领域尚不具备赋予其主体地位的条件,将其确定为一种人工智能服务更加适宜。其次在侵权责任的认定路径上,宜将人工智能服务提供者划为过错责任原则,并依据侵权主体各自过失承担相应的责任,同时优化人工智能算法实现可解释性。鉴于生成式人工智能的侵权风险复杂性,应从侵权责任主体明晰、隐私保护、算法透明及生成内容监管等方面进行规避。

The tort liability of generative artificial intelligence represented by ChatGPT should be explored from the legal positioning of artificial intelligence itself and the determination of tort liability of related subjects.At present,the field of generative artificial intelligence infringement liability represented by ChatGPT does not yet have the conditions to grant it a subject status,and it is more appropriate to define it as an artificial intelligence service.Secondly,in the determination of infringement liability,it is advisable to classify artificial intelligence service providers as the principle of fault liability,and to bear corresponding responsibilities based on the respective faults of the infringing subjects,while optimizing artificial intelligence algorithms to achieve interpretability.In view of the complexity of the infringement risk of generative artificial intelligence,it should be avoided from the aspects of clear infringement liability subject,privacy protection,algorithm transparency and generated content supervision.