The Structure-Aware Temporal Graph RAG (SAT-Graph RAG) addresses core limitations of standard Retrieval-Augmented Generation in the legal domain by providing a verifiable knowledge graph that models hierarchical structure, temporal evolution, and causal events of legal norms. However, a critical gap remains: how to reliably query this structured knowledge without sacrificing its deterministic properties. This paper introduces the SAT-Graph API, a formal query execution layer centered on canonical actions-atomic, composable, and auditable primitives that isolate probabilistic discovery from deterministic retrieval. These actions enable: (i) high-precision hybrid search; (ii) robust reference resolution; (iii) point-in-time version retrieval; and (iv) auditable causal tracing. We demonstrate how planner-guided agents can decompose complex queries into Directed Acyclic Graphs (DAGs) of these actions. This two-layer architecture transforms retrieval from an opaque black box to a transparent, auditable process, directly addressing Explainable AI (XAI) requirements for high-stakes domains.

NOR : HRUX2527915D JORF n°0239 du 11 octobre 2025 Texte n° 1 Le Président de la République, Vu l'article 8 de la Constitution, Décrète : Article 1 M. Sébastien LECORNU est nommé Premier ministre. Article 2 Le présent décret sera publié au Journal officiel de la République française. Fait le 10 octobre 2025. Emmanuel Macron

Décret du 10 octobre 2025 portant nomination du Premier ministre

NOR : HRUX2527313D JORF n°0235 du 7 octobre 2025 Texte n° 2 Le Président de la République, Vu l'article 8 de la Constitution ; Vu le décret du 9 septembre 2025 portant nomination du Premier ministre ; Vu le décret du 5 octobre 2025 relatif à la composition du Gouvernement ; Vu le décret du 6 octobre 2025 relatif à la cessation des fonctions du Gouvernement ;

NOR : HRUX2527269D JORF n°0235 du 7 octobre 2025 Texte n° 1 Le Président de la République, Vu la Constitution, notamment l'article 8 ; Vu la lettre, remise le 6 octobre 2025, par laquelle le Premier ministre a présenté au Président de la République la démission du Gouvernement, Décrète : Article 1 Il est mis fin, sur la présentation de la démission du Gouvernement, aux fonctions de M.

NOR : HRUX2527171D JORF n°0234 du 5 octobre 2025 Texte n° 70 Le Président de la République, Vu l'article 8 de la Constitution ; Vu le décret du 9 septembre 2025 portant nomination du Premier ministre ; Sur proposition du Premier ministre, chargé de la planification écologique et énergétique, Décrète : Article 1 Sont nommés ministres : Mme Elisabeth BORNE, ministre d'Etat, ministre de l'éducation nationale, de l'enseignement supérieur et de la recherche ;

This study addresses the limitations of traditional keyword-based legal information systems by proposing a robust model for managing legislative changes. We introduce the Legal-Onto model, integrated with Vietnamese SBert, to enhance the semantic representation and accessibility of legal documents. The model leverages an ontology-based framework to represent legal structures and employs a Sentence-BERT variant tailored for Vietnamese to compare legal texts at multiple granularities–chapters, sections, articles, clauses, and points. Unlike prior methods that rely solely on the highest similarity score, our approach retains multiple high-similarity matches to mitigate misalignment caused by overlapping legal provisions. This strategy ensures accurate mapping of new legislative elements and prevents erroneous prioritization. The model architecture integrates relational and conceptual graphs with natural language processing techniques to systematically detect and highlight legislative updates. Experimental evaluations on the Vietnamese Land Laws of 2013 and 2024 demonstrate the model’s effectiveness in identifying both structural and content changes. Results show that the Legal-Onto model significantly improves legal knowledge management by achieving high alignment accuracy and reducing manual analysis efforts.

Navigating and retrieving relevant excerpts of legislation is challenging, requiring time and effort, especially to fine-tune appropriate input search queries. Furthermore, the continuously growing, heterogeneous body of laws, combined with a deep interconnection among normative acts, adds a layer of complexity: some potentially relevant rules may be hidden in articles that, through multiple citations and references, might be relevant for the input query. Traditional search systems, based on keywords or more sophisticated approaches as BM25 or TF-IDF, do not support such flexible exploration, being ineffective at handling contextual information. To address these challenges, recent research proposed using graph data models for legislative knowledge management, introducing a straightforward approach to handling network complexity. They adopted the Property Graph data structure, demonstrating how it provides semantics and navigation power, supporting advanced querying tools for legislative acts, and implemented it on the Italian legislation. In this paper, we build on recent results on legislative knowledge management with graphs by proposing LegisSearch, an effective navigation system that, combining the graph data model with pre-trained Large Language Models and universal text embeddings, allows users to conduct powerful searches within a legislative system. We implement LegisSearch within the Italian graph of national laws, and we test its performance across multiple domains by comparing its search results with those provided in specific thematic areas by Italian ministries on their official websites, demonstrating its superior retrieval performance over traditional search systems and testing the contribution of each component.

Artificial Intelligence (AI) drives legal innovation. By integrating advanced machine learning, natural language processing, and large language models, AI achie

The Commission's activities to safeguard European democracy, promote free and fair elections and uphold electoral rights of EU citizens.

Critics of generative AI often describe it as a “plagiarism machine.” They may be right, though not in the sense they mean. With rare exceptions, generative AI

Aujourd'hui, la Commission a défini deux stratégies pour faire en sorte que l'Europe exploite pleinement le potentiel de l'IA, en favorisant l'adoption dans des secteurs clés et en plaçant l'Europe à l'avant-garde de la science fondée sur l'IA.

Wikipedia is one of the most widely accessed sources of information worldwide, playing a crucial role in global knowledge dissemination across languages and cul...

Service Public fête ses 25 ans avec une nouvelle URL et une nouvelle identité graphique

Mise à jour 02h17 du matin, le mardi 7 octobre 2025… alors que le JO du 7 octobre vient d’être publié 1/ Est-ce le Gouvernement le plus court, en nombre de jours de plein exercice, de l…

Legal codes have existed since the time of Hamourabi’s Code. Ancient codes enumerated customs. Modern codes, like the Napoleonic Code, were new points of depart

Milburn-Knizner, April and Milburn-Knizner, April and Montcalm, Megan E.K., TL;DR: The Truth About Gen Z Law Students & Why They Need More, Not Less (August 01, 2025). Available at SSRN: https://ssrn.com/abstract=5471486 or http://dx.doi.org/10.2139/ssrn.5471486

This piece critically analyzes whether AI-based legislative drafting jeopardizes democratic legitimacy in South Asia's plural constitutional orders, and is spec

Why edit a constitutional law casebook? One might want to shape how professors teach the subject and how students learn it. But the influence of casebook editor

Retrouvez, dans le numéro de septembre 2025 de la lettre de la justice administrative, toute l’actualité jurisprudentielle et consultative du Conseil d’État.