VidhiAI: An Explainable Hybrid Architecture for Legal Document Summarization and Regulatory Compliance in Indian Jurisprudence

Artificial Intelligence is revolutionizing the legal industry by enhancing the efficiency of legal research, documentation, and compliance verification. This work introduces VidhiAI, a domain-specific large language model (LLM) tailored for Indian jurisprudence. The novelty of VidhiAI lies in its hybrid architecture, which combines a fine-tuned LLM with Retrieval-Augmented Generation (RAG) to perform accurate, context-aware legal document summarization and statutory compliance checks. Optimized on curated Indian legal datasets, VidhiAI generates structured summaries and cross-references statutory requirements to ensure jurisdictional compliance. We present the model architecture, data curation process, and a comparative evaluation against existing legal AI tools. The study also addresses challenges in legal AI, including data bias, ethical concerns, and regulatory constraints, positioning VidhiAI as a step toward intelligent legal assistants that support legal professionals while ensuring legal integrity.