中圖分類號(hào)：D922.16文獻(xiàn)標(biāo)識(shí)碼：ADOI:10.19358/j.issn.2097-1788.2025.04.013
引用格式：陳鏡霖. 司法自動(dòng)化決策的紓困原理與算法解釋權(quán)路徑展開［J］.網(wǎng)絡(luò)安全與數(shù)據(jù)治理,，2025,，44（4）：84-90.

The principle of risk mitigation in judicial automated decision-making and the path of algorithmic explainability

Abstract： The application of judicial automation decision-making can meet the purpose of judicial prevention, and improve the integration process of case information,so it has certain legitimacy and feasibility. However, the current application of judicial automation decision-making still faces many obstacles: at the procedural level, algorithmic black boxes lead to decision-making logic that cannot be understood by the public; At the entity level, excessive involvement of algorithms can easily trigger the risk of improper liability deviation. Therefore, legal frameworks should ground judicial accountability in the GDPR′s and China′s Personal Information Protection Law′s right to algorithmic explanation. This enables comprehensive remedies across pre-decision, ongoing, and post-decision phases while the inherent shortcomings of black box judicial decision-making should be eliminated.These measures collectively clarify ownership structures and operational protocols for algorithmic interpretation, thereby institutionalizing the partial transfer of judicial authority to automated systems.Concretely,it is necessary to improve the subject,object,and behavioral requirements of judicial algorithm interpretation power.Procedural mechanisms for automated judicial decisions must be standardized, while the responsibility mechanism of judicial automation decision-making should be developed.By clarifying the ownership and operation mode of algorithm interpretation power in judicial automation decision-making, institutional guarantees are provided for the partial transfer of judicial power to algorithms.

Key words : judicial automated decision-making; algorithmic explainability; algorithmic black box; accountability mechanism