ChatGPT技术带来的新法律风险及应对措施  

New Legal Risks and Countermeasures Brought by ChatGPT Technology

作　　者：杨琳 杨昂 Yang Lin;Yang Ang(School of Intellectual Property,Dalian University of Technology,Dalian,China;School of Shipping Economics and Management,Dalian Maritime University,Dalian,China)

机构地区：[1]大连理工大学知识产权学院,辽宁大连116024 [2]大连海事大学航运经济与管理学院,辽宁大连116026

出　　处：《社会科学论坛》2025年第2期141-155,共15页Tribune of Social Sciences

基　　金：教育部人文社会科学研究青年基金项目《制度型开放背景下专利诉讼保护的企业创新效应及优化路径》阶段性成果,项目编号:23YJCZH263;教育部人文社会科学研究青年基金项目《面向条件差异化的港口智慧化转型多情境投资组合优化研究》阶段性成果,项目编号:22YJCZH210;国家自然科学基金青年科学基金项目《“双碳”目标下智慧港口综合能源系统低碳转型机理分析与政策优化研究》阶段性成果,项目编号:72203029。

摘　　要：自聊天机器人ChatGPT问世以来,独特的连续对话式创作方法不仅赋予ChatGPT高度的创造能力,也带来了不同于传统人工智能系统的新法律风险。新法律风险可以分为三方面:一是可以被用以拼凑和还原的碎片化的敏感信息,强化了敏感信息的泄露风险;二是改变了使用者与人工智能系统间关系的底层逻辑,造成著作权人困境;三是可能形成定向反馈机制,显著提高伦理规范风险。对此,可以从提高使用透明度、健全入口端监管机制,明确使用者双重身份,建立反馈机制和社会参与机制等方面入手,降低ChatGPT的法律风险,提高其服务人类的安全性和水准。Since the emergence of chatbot ChatGPT,the unique continuous conversational creation method not only gives ChatGPT a high degree of creative ability,but also brings new legal risks that are different from traditional artificial intelligence systems.The new legal risks can be divided into three aspects:first,it can be used to piece together and restore fragmented sensitive information,strengthening the risk of leakage of sensitive information;second,the underlying logic of the relationship between users and artificial intelligence systems creates a dilemma for copyright owners;third,a directional feedback mechanism may be formed to significantly increase the risk of ethical regulations.In this regard,we can start from improving the transparency of use,improving the portal supervision mechanism,clarifying the dual identities of users,and establishing feedback mechanisms and social participation mechanisms,etc.,to reduce the legal risks of ChatGPT and improve the security and level of its service to mankind.

关 键 词：人工智能 法律风险 版权 个人信息 伦理风险 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]