Agentic Retrieval-Augmented Generation for Verifiable Regulatory Compliance in Urban Cyber-Physical Systems
Abstract
Abstract Urban Cyber-Physical Systems (UCPS) support sustainable cities through the integration of Internet of Things (IoT), Artificial Intelligence (AI), and digi-tal infrastructures for environmental monitoring, urban planning, and resilient public services. However, ensuring regulatory compliance is increasingly chal-lenging because environmental, cybersecurity, municipal planning, and AI gov-ernance regulations evolve continuously across jurisdictions. Existing Retrieval-Augmented Generation (RAG) approaches improve evidence retrieval but remain vulnerable to hallucinations, regulatory supersession errors, and jurisdictional inconsistencies due to the absence of deterministic regulatory verification. This paper proposes a Verification-First Agentic Retrieval-Augmented Generation framework for trustworthy regulatory compliance reasoning in UCPS. The framework incorporates a Retrieval Agent, a deterministic Critic Agent, and a Formatting Agent to validate temporal validity, regulatory supersession, juris-dictional applicability, and compliance scope before language-model generation. A curated regulatory corpus and a benchmark comprising 78 regulatory com-pliance probes were used for evaluation against BM25, Dense RAG, Self-RAG, and Corrective RAG (CRAG). The proposed framework achieved a superses-sion detection rate of 0.88, perfect standalone Critic Agent validation (Precision = Recall = F1-score = 1.00), improved retrieval ranking quality (nDCG@5 = 0.75), and competitive…
Read the paper
Related studies
- Knowledge Management Perspective of Generative Artificial Intelligence
- Navigating Generative Artificial Intelligence Promises and Perils for Knowledge and Creative Work
- Characterizing the evolving research on enterprise content management
- Large language models and multimodal foundation models for precision oncology
- From Generative AI to Extended Reality: Multidisciplinary Perspectives on the Challenges, Opportunities, and Future of Educational Computing
- Generative Artificial Intelligence: A New Engine for Advancing Environmental Science and Engineering
Go deeper in Evidence Hub
Evidence Hub adds a study data sheet (design, sample size, evidence level, key numbers), an AI summary and a chat grounded in the paper's text.