算法行政的风险场景及其法律规制——基于算法法权关系架构的分析

The Risk Scenarios of Algorithmic Administration and Its Legal Regulation-Analysis Based on Relational Architecture of Algorithmic Legal Rights

算法行政具有客观、高效优势,但同时伴随着技术内生的不确定性风险、法律层面权力行使与权利保障的挑战,以及价值层面技术理性吞噬价值理性等桎梏,亟需法律破解。规制对象是法律制度构建之前提,将“算法法权”作为算法行政风险的法律规制对象具有合理性和可行性,通过“权力—权利”的法权结构,可实现算法行政中“私权利、公权力、私权力”的整体性规制。质言之,算法行政风险的法律规制需要以“算法法权”的科学配置为核心,建构法律、伦理、技术与自我规制一体化的规制体系。

In the era of algorithmic advancement,in response to the transformational demands of digital governance,governments have promoted the digital transformation of government based on the concept of“government as a platform”,laying a foundation for the operation of algorithmic administration.Algorithmic trust and data advantages can greatly mitigate the bounded rationality of human decision-making.The algorithmic operation of law can facilitate the standardized,rigorous,trustworthy,and efficient execution of legal rules,particularly in the analysis of massive data and the enforcement of non-site-based law enforcement so as to achieve consistency,immediacy,and universality in administrative decision-making.However,algorithmic administration is faced with various risks and challenges at the technological,legal,and value levels,which may substantially affect public interests and the individual rights of administrative counterparts.As a result,legal responses become necessary.Existing studies show that algorithmic has changed the traditional“power-right”framework,leading to the decentralization of power subjects,the expansion of power scope,and bidirectional power interactions.Notably,these analyses only focus on algorithmic power and algorithmic rights but lack systematic regulation research on the interrelationships among“algorithmic public power-algorithmic private power-algorithmic private rights”.Therefore,this paper introduces the“legal right”theory into the legal regulation of algorithmic administrative risks and we propose the concept of“algorithmic legal rights”and aim to achieve structured and comprehensive regulation of“algorithmic private rights,algorithmic public power,and algorithmic private power”in algorithmic administration.By establishing the relationships of algorithmic legal rights,namely“algorithmic public power-algorithmic private power”,“algorithmic private power-algorithmic private rights”,and“algorithmic public power-algorithmic private rights”,a systematic regulation of algorithmic administrative risks can be achieved,which can strengthen the protection of algorithmic private rights for administrative counterparts and restrain the exercise of algorithmic public power.This framework can provide a feasible approach for the iterative development and risk adjustment of algorithmic administration.In summary,the legal regulation of algorithmic administrative risks requires focusing on the scientific configuration of“algorithmic legal rights”and on constructing an integrated regulatory system that combines law,ethics,technology and self-regulation.In terms of the design’s approach,it includes constructing norms to address uncertainties of algorithmic administration from the dimensions of algorithmic uncertainty risk,cognitive uncertainty risk,and institutional uncertainty risk;promoting coordinated regulation of different relationships among government public power,platform private power,and algorithmic private rights,with the architecture of algorithmic legal right relationships at its core;reconciling the deviation trend between technological rationality and value rationality,that algorithmic administrative regulation should uphold the value criteria of individual justice and digital justice.