摘要
《个人信息保护法》第24条标志着算法解释权在制度层面得以确立。然而,算法控制者往往将算法视为核心竞争力并以商业秘密的形式进行保护,受算法影响者对算法解释的合理诉求与算法控制者对算法保密的现实需要之间不可避免地存在着冲突与张力。面对二者的紧张关系,无论是全然废除算法解释规则,还是算法解释权当然优先,抑或诉诸漫无边际的利益衡量,均非可取之道。为避免冲突激化,不宜将“算法黑箱”完全打开,而只需将其“掀开最小缝隙”,至受其不利影响者可见的程度即可。在“掀开最小缝隙”理论下,算法解释权的行使前提“对个人权益有重大影响”应从严把握,算法解释的内容应限定为算法运行逻辑而非算法本身。同时,商业秘密的“秘密性”需作出澄清,受算法影响者还应负有初步证明责任与保密协议的签订义务。如此方可在最大程度上实现算法解释与商业秘密的冲突化解。
Article 24 of the Personal Information Protection Law means the establishment of the right of algorithm interpretation at the institutional level.However,algorithm controllers often regard algorithms as the core competitiveness of enterprises and protect them in the form of trade secrets.There is an inevitable conflict and tension between the reasonable demands for algorithm interpretation of individuals affected by the algorithm and the realistic needs for algorithm secrecy of algorithm controllers.In the face of the tension between the two,completely abolishing the rules of algorithm interpretation,establishing the priority of the right of algorithm interpretation,or resorting to the rambling measure of interests is all wrong.In order to avoid the intensification of conflicts,it is not appropriate to fully open the“black box”of algorithm,but only to“open a gap”to the extent that the users who are adversely affected by it can see it.Under such theory,the premise of exercising the right of algorithm interpretation,which is“having a significant impact on personal rights and interests”,should be strictly grasped,and the content of algorithm interpretation should be limited to the operation logic of the algorithm rather than the algorithm itself.Meanwhile,the“secrecy”of trade secrets need to be clarified,and the individuals affected by the algorithm should also bear the initial burden of proof and the obligation to sign a confidentiality agreement.Only in this way can the conflict between algorithm interpretation and trade secret be best resolved.
作者
刘琳
LIU Lin(Law School of Nanjing University,Nanjing 210093)
出处
《行政法学研究》
北大核心
2023年第2期168-176,共9页
ADMINISTRATIVE LAW REVIEW
基金
2021年国家社科基金项目“民法典中隐私与个人信息的二元保护模式研究”(项目编号:21BFX084)
中国法学会2022年度部级课题“人工智能算法解释规则与商业秘密制度的冲突与协调”(项目编号:CLS(2022)D104)
2021年安徽省高校人文社会科学研究重点项目“商业秘密的民刑救济路径之独立与衔接问题研究”(项目编号:SK2021A0002)。
关键词
算法解释权
商业秘密
算法权力
算法歧视
算法黑箱
Algorithm Interpretation Rights
Trade Secret
Algorithmic Power
Algorithmic Discrimination
Black Box of Algorithmic