风险预防:AIGC视域下个人信息保护的模式革新

Risk Prevention:Innovation of Personal Information Protection Model from the Perspective of AIGC

随着“人工智能+”行动的开展,以ChatGPT、Sora、文心一言为代表的AIGC将逐步实现商业化落地,然其亦伴随着多重信息安全风险。基于AIGC的技术原理,可将这些风险归纳为“三阶段风险”:在准备阶段,其将弱化告知同意原则,侵犯信息主体隐私权;在运行阶段,其将虚置目的限制原则,加大信息泄露风险;在输出阶段,其可能生成虚假性与歧视性等不良信息内容。传统个人信息保护模式难以应对AIGC带来的信息安全风险与挑战。而以“风险预防”为导向的个人信息保护模式,既有助于平衡数字经济发展与个人信息保护,也有利于合理分配信息安全风险,实现个人信息的高效保护。鉴于此,应将“风险预防”贯彻于AIGC“三阶段风险”之中,即推动“告知同意”从“个人到风险”转移,促使“目的限制”向“合理必要”转化,构建不良信息内容全过程预防机制,以促进个人信息保护与新质生产力的协同发展。

With the implementation of the“AI+”initiative,AIGC,represented by ChatGPT,Sora and ERNIE Bot,will gradually realize its commercialization,but it is also accompanied by multiple information security risks.Based on the technical principles of AIGC,these risks can be summarized as“three-stage risks”:in the preparation stage,they will weaken the principle of informed consent and infringe on the privacy rights of information subjects;During the operation phase,it will adopt the principle of virtual purpose limitation to increase the risk of information leakage;In the output stage,it may generate false and discriminatory information content.The traditional personal information protection model is difficult to cope with the information security risks and challenges brought by AIGC.The“risk prevention”oriented personal information protection model not only helps balance the development of the digital economy and personal information protection,but also facilitates the rational allocation of information security risks and achieves efficient protection of personal information.In view of this,“risk prevention”should be implemented in the“three-stage risk”of AIGC,that is,to promote the transfer of“informed consent”from“individual to risk”,to promote the transformation of“purpose restriction”to“reasonable and necessary”,and to construct a whole process prevention mechanism for harmful information content,in order to promote the coordinated development of personal information protection and new quality productivity.