元宇宙时代下人工智能法律风险与犯罪论体系探究  

Legal Risks and Crime Theory System with Artificial Intelligence in the Metaverse Era

作　　者：王学光[1] WANG Xue-guang(Criminal Justice College,East China University of Political Science and Law,Shanghai 200042,China)

机构地区：[1]华东政法大学刑事法学院,上海200042

出　　处：《南通大学学报（社会科学版）》2023年第6期23-34,共12页Journal of Nantong University:Social Sciences Edition

摘　　要：元宇宙是由电子计算机、虚拟现实和量子技术等制成,介由高度发达的脑机接口技术,提供完全潜行环境的平行世界。在元宇宙环境中,管理员和强人工智能的共同犯罪即将成为一种新的犯罪趋势,为了解决犯罪主体的不适应性,有必要对我国通说温和的三阶层犯罪要件构成说再探讨。对于构成要件符合性,其实质就是管理员与人工智能监管程序类型化的违法性行为。为了解决元宇宙中精神损害所指向的实行行为无法被违法性所涵摄的问题,违法性的实质应当从泷川幸辰和前田雅英的实质利益损害的观点转向麦耶所主张的伦理规范违反说,而伦理违反性也应当理解为违法性更深层次的概念。因此从伦理规范违反说的角度应当对人工智能监管程序的程序犯意、危害行为、行为结果认定其违法性,并从危害行为、行为结果的角度认定系统管理员行为的违法性,但共同犯罪的认定范围仍限定于双方共同的行为部分。对于违法阻却事由,通过对元宇宙系统资源的财产属性的探讨,可明确用户的自力救济行为的相应刑法学属性。而对于责任要素,其自身情感和价值判断的脆弱性决定了刑罚所想表达的痛苦并不能如愿传达给强人工智能,因此对于强人工智能而言,有责性的本质应当转化为对侵害法益、伦理行为的预告性制裁。若具有相对意志自由的人工智能监管程序在对刑法所保护的权益进行充分衡量后仍实施不法行为,则应认定其犯罪故意,确定其相应的刑事责任。The metaverse is a parallel world made of electronic computers and quantum technology,mediated by highly developed brain-computer interface technology,providing a fully submerged environment.The joint crime of administrators and strong artificial intelligence is about to become a new criminal trend in the meta cosmic environment.In order to solve the inadaptability of the criminal subject,it is necessary to re-explore the generally mild theory of three levels of crime elements in China.The essence of constituent element compliance is the illegal behavior of administrators and AI supervision procedures.In order to solve the problem that the practice of mental damage in the meta universe cannot be covered by illegality,the essence of illegality should be changed from the viewpoint of material interest damage of Yuchen Takigawa and Yaying Maeda to the theory of violation of ethical norms advocated by Meyer,and ethical violation should also be understood as a deeper-layer concept of illegality.Therefore,from the perspective of the theory of violation of ethical norms,we should recognize the illegality of the procedural intent,harmful behavior,and behavior results of the AI supervision procedure,and recognize the illegality of the system administrator’s behavior from the perspective of harmful behavior and behavior results.However,the scope of recognition of joint crime is still limited to the common behavior of both parties.As for the causes of illegal refusal,through the discussion of the property attribute of the resources of the meta universe system,the corresponding criminal law attribute of the user’s self-help behavior can be clarified.As for the element of responsibility,the vulnerability of its own feelings and value judgments determines that the pain that the penalty wants to express cannot be conveyed to strong AI as expected.Therefore,for strong AI,the nature of responsibility should be transformed into a prescriptive sanction against violations of legal interests and ethical behaviors.If the AI superv

关 键 词：元宇宙 人工智能 犯罪论体系 法教义学 

分 类 号：D923.41[政治法律—民商法学] TP18[政治法律—法学]