Navigating Global AI Regulation: A Multi-Jurisdictional Retrieval-Augmented Generation System

TL;DR

This paper introduces a retrieval-augmented system for navigating global AI regulations, effectively handling diverse legal documents across multiple jurisdictions with high accuracy.

Contribution

It presents novel techniques for chunking, retrieval routing, and re-ranking tailored to legal texts, improving multi-jurisdictional AI regulation navigation.

Findings

Achieved 0.87 faithfulness and 0.84 relevancy on 50 queries.

Single-entity queries reach 0.86 faithfulness and 0.92 relevancy.

Multi-jurisdictional comparison queries reach 0.88 faithfulness and 0.75 relevancy.

Abstract

Navigating AI regulation across jurisdictions is increasingly difficult for policymakers, legal professionals, and researchers. To address this, we present a multi-jurisdictional Retrieval-Augmented Generation system for global AI regulation. Our corpus includes 242 documents across 68 jurisdictions, ranging from formal legislation like the EU AI Act to unstructured policy documents such as national AI strategies. The system makes three technical contributions: type-specific chunking that preserve legal structure across heterogenous documents; conditional retrieval routing with entity detection and metadata for legal citations; and priority-based re-ranking to boost enacted legislation over policy and secondary sources. Evaluation of 50 queries reveals strong performance across both single-entity and multi-jurisdictional questions, achieving 0.87 average faithfulness and 0.84 average answer relevancy. Single-entity queries achieve 0.86 average faithfulness and 0.92 average answer relevancy, while multi-jurisdictional comparison queries achieve 0.88 average faithfulness and 0.75 average answer relevancy. These findings highlight the effectiveness of domain-specific retrieval strategies for navigating complex, heterogenous regulatory corpora.