The Assumption and Limitation of Indirect Infringement Liability of Generative Artificial Intelligence Service Provider
Generative artificial intelligence represented by ChatGPT does not have the qualification of the persons of the civil law,the service provider should bear the tort liability if the generative artificial intelligence cau-ses damage to others.The service provider with the dual identity of content producer and platform manager has the duty of maintaining safety,such as auditing.If the service provider violates the duty of maintaining safety,the corresponding supplementary liability with the attribute of indirect tort liability should be borne in principle and the service provider can claim right of indemnification.However,the service provider should bear joint responsibility and even punitive damages under the circumstance of"knows or should have known".If infringing contents derive from illegal resources,are generated illegally or inputting information il-legally,and are generated illegally with the service user propagating them to cause damages to the third par-ty,the service provider will be prohibited from claiming indemnification due to the violation of the dual audi-ting obligation.Meanwhile,the liability exemption rules for the application of the"safe harbor rules"should be established in a typological manner,the determining mechanism of grounds for elimination of illegality for the selective application of the rule of fair use of copyright should be shaped and the liability exemption sys-tem for the limited application of the rule of fair use of personal information should be constructed in order to systematically establish the limitation rules of the liability for indirect infringement of the service provider.
Generative Artificial Intelligence Service ProviderQualification of SubjectTortObliga-tion AssumptionObligation Limitation
NETL
NSTL
万方数据
