论深度伪造技术侵犯他人声音权益的民法保护

On Civil Law Protection of Infringement of Others’ Voice Rights and Interests by DeepFakes Technology

摘要: 深度伪造是一种人工智能的新兴技术，通过“深度学习”和“伪造”两个基本系统可以实现对他人声音的模仿，达到以假乱真的程度，引发了不法分子利用深度伪造技术侵犯他人声音人格权益现象的发生；而声音信息作为个人信息的一种，未经许可使用他人声音数据亦会侵犯权益人的个人信息权。与传统侵权行为相比，深度伪造侵权存在着亟待解决的新生问题，具体包括技术提供者的责任难以界定、受害人举证困难、侵权人身份难以确定三大内容。在明确声音权益保护法律依据的基础上，通过对技术提供者、音频传播者课以相应义务和完善深度伪造行为人责任追究规则，实现声音权益的妥善保护。

Abstract: DeepFakes is a new technology of artificial intelligence. Through the two basic systems of “deep learning” and “fakes”, one can imitate the voice of others to the extent that it is false, which leads to the phenomenon that criminals use DeepFakes technology to infringe the personality rights and interests of others’ voices. As a kind of personal information, the use of other people’s voice data without permission will also infringe the rights of personal information of the obligee. Compared with the traditional infringement, the DeepFakes infringement has new prob-lems that need to be solved urgently, including the difficulty in defining the responsibility of the technology provider, the difficulty in providing evidence for the victim, and the difficulty in deter-mining the identity of the infringer. On the basis of clarifying the legal basis for the protection of voice rights and interests, the voice rights and interests can be properly protected by imposing corresponding obligations on technology providers and audio communicators and improving the rules for accountability of perpetrators of DeepFakes.