ChatGPT类生成式人工智能犯罪风险及其治理路径探究  

An Exploration of Criminal Risks Posed by ChatGPT-like Generative Artificial Intelligence and Their Governance Pathways

作　　者：王文松 张郁 Wang Wensong;Zhang Yu(Gansu University of Political Science and Law,Public Security Institute,Gansu,Lanzhou,730070)

机构地区：[1]甘肃政法大学公安分院,甘肃兰州730070

出　　处：《科技智囊》2024年第11期43-52,共10页Think Tank of Science & Technology

基　　金：国家级大学生创新创业训练计划项目“粤港澳大湾区刑事司法互助机制研究”(202311406004);省级大学生创新创业训练计划项目“互联网平台算法推荐司法治理研究”(202411406057);2023年度甘肃省社科规划项目“甘肃省涉众型经济犯罪治理路径研究”(2023YB117)。

摘　　要：[研究目的]随着ChatGPT类生成式人工智能技术的迅猛发展,其可能带来的隐私泄露与数据安全风险、侵犯知识产权犯罪风险、经济与金融犯罪风险、社会信任与信息真实性风险等诸多问题日益凸显,进而催生了治理生成式人工智能犯罪风险的时代诉求。探究一套系统的治理框架,以应对和减轻此类风险,确保生成式人工智能技术的合理利用与可持续发展有重要意义。[研究方法]运用案例分析法,深入剖析了具体案例中生成式人工智能犯罪的表现形式、成因及危害;同时,采用了比较分析法,对比了国内外在治理生成式人工智能犯罪风险方面的法律法规、政策、举措及实践经验。[研究结论]在当前生成式人工智能犯罪风险的治理体系中,存在法律规制的滞后性、技术中立原则界限的模糊、技术监测与识别能力的限制、跨境监管合作的缺失等问题,亟待从完善法律法规体系、强化伦理准则与行业自律、提升技术监管与防御能力、加强国际合作与信息共享、提升公众意识与教育等多维度综合治理,从而实现生成式人工智能的健康发展。[Research purpose]With the rapid development of generative aritficial intelligence exemplified by ChatGPT,numerous issues have increasingly come to the fore,including risks of privacy breaches and data security,infringement of intellectual property rights,economci and financial crimes,and threats to socia l trust and information authenticity.Consequently,there arises an urgent need for an era of governance to address the criminal risks associated with generative AI.This study endeavors to explore a comprehensive governance framework aimed at addressing and mitigating such risks,thereby ensuring the rational utilization and sustainable development of generative AI technologies.[Research method]Utilizing the case analysis method,this study delves into the manifestations,causes,and harms of generative artificial intelligence-related crimes in specific cases.Additionally,it employs a comparative analysis approach to contrast the legal frameworks,policies,measures,and practical experiences in managing the risks of generative AI-related crimes both domestically and internationally.[Research conclusion]In the current governance framework addressing criminal risks associated with generative artificial intelligence,several issues persist,including the lag in legal regulation,the ambiguity surrounding the boundaries of the principle of technological neutrality,limitations in technical monitoring and identification capabilities,and the lack of cross-border regulatory cooperation.To realize the healthy development of generative artificial intelligence,it is imperative to adopt a multi-faceted and comprehensive approach,encompassing the refinement of legal and regulatory frameworks,the reinforcement of ethical guidelines and industry self-discipline,the enhancement of technical monitoring and defense capabilities,the strengthening of international cooperation and information sharing,and the eleavtion of public awareness and education.

关 键 词：ChatGPT 生成式人工智能 犯罪风险 治理困境 治理路径 

分 类 号：D917.6[政治法律—法学]