Education, Science, Technology, Innovation and Life
Open Access
Sign In

Risks and Mitigation Strategies of Artificial Intelligence in Judicial Adjudication: A Human-AI Collaborative Framework for Judicial Reasoning

Download as PDF

DOI: 10.23977/law.2025.040220 | Downloads: 7 | Views: 399

Author(s)

Li Dan 1

Affiliation(s)

1 Faculty of Law, Beijing Normal University, Beijing, 100000, China

Corresponding Author

Li Dan

ABSTRACT

Artificial Intelligence (hereinafter referred to as AI) is being rapidly integrated into judicial adjudication, with big data-driven technologies playing a pivotal role. The transformation underpinning intelligent adjudication-a shift from text-based to code-driven judicial logic-has not only catalyzed the emergence of AI jurisprudence but also reshaped the concept of justice centered on the consistent adjudication of similar cases. However, the coexistence of complexity and opacity in AI-powered judicial systems may undermine their purported advantages. Beyond the black-box effects arising from the irreducibility of AI-driven judicial processes, issues persist in data and algorithms, two critical components of the AI judicial ecosystem. Consequently, judges' interpretive responsibilities now encompass new dimensions: verifying and integrating the foundational conditions of AI-assisted adjudication. To align AI technologies with the objectives of human judicial decision-making, judges should focus on four hermeneutic checkpoints through human-machine collaboration: (1) Is the case dataset sufficiently comprehensive? (2) Is the adjudicative outcome normatively justified? (3) Is the causal relationship appropriately established? (4) Is the symmetry between affirmative and negative computational reasoning maintained.

KEYWORDS

Artificial Intelligence; Judicial Adjudication; Human-Machine Collaboration; Judge's Interpretation

CITE THIS PAPER

Li Dan. Risks and Mitigation Strategies of Artificial Intelligence in Judicial Adjudication: A Human-AI Collaborative Framework for Judicial Reasoning. Science of Law Journal (2025) Vol. 4: 161-170. DOI: http://dx.doi.org/DOI: 10.23977/law.2025.040220.

REFERENCES

[1] Gadamer, H. G. (2013). Hermeneutics II: Truth and method (H. Hong, Trans.). Beijing: The Commercial Press. (Original work published 1960) 
[2] Huanqiu.com. (2016, May 16). IBM develops the world's first AI lawyer-Ross. Retrieved March 1, 2018, from http://tech.huanqiu.com/original/2016-05/8935344.html
[3] Luo, W. P. (2018). The problem induction and prospect of AI adjudication. Journal of National Prosecutors College, 5, 16-31. 
[4] Yan, J. Y. (2017, December 3). Shanghai's intelligent case-handling system covers civil, commercial, and administrative fields. People's Court Daily, p. 001.

All published work is licensed under a Creative Commons Attribution 4.0 International License.

Copyright © 2016 - 2031 Clausius Scientific Press Inc. All Rights Reserved.