RAG System for Supporting Japanese Litigation Procedures: Faithful Response Generation Complying with Legal Norms
Yuya Ishihara, Atsushi Keyaki, Hiroaki Yamada, Ryutaro Ohara, Mihoko Sumida
TL;DR
The paper tackles the challenge of applying Retrieval-Augmented Generation to Japanese medical litigation while enforcing strict legal-norm compliance. It outlines a norm-compliant RAG framework anchored on three pillars: controlling knowledge sources in line with procedural requirements, ensuring attribution and faithfulness to retrieved context, and using time-appropriate sources. It introduces concepts like Data Attribution and Faithfulness Score to evaluate source–response alignment and discusses time-stamped sourcing to address issue-specific knowledge validity. The work aims to enable reliable, legally compliant AI assistance for judges and experts in civil medical cases, with future work on formal methods and experiments.
Abstract
This study discusses the essential components that a Retrieval-Augmented Generation (RAG)-based LLM system should possess in order to support Japanese medical litigation procedures complying with legal norms. In litigation, expert commissioners, such as physicians, architects, accountants, and engineers, provide specialized knowledge to help judges clarify points of dispute. When considering the substitution of these expert roles with a RAG-based LLM system, the constraint of strict adherence to legal norms is imposed. Specifically, three requirements arise: (1) the retrieval module must retrieve appropriate external knowledge relevant to the disputed issues in accordance with the principle prohibiting the use of private knowledge, (2) the responses generated must originate from the context provided by the RAG and remain faithful to that context, and (3) the retrieval module must reference external knowledge with appropriate timestamps corresponding to the issues at hand. This paper discusses the design of a RAG-based LLM system that satisfies these requirements.