Publication

Abstract : The need for effective solutions in reviewing, risk identification, and compliance management of complex legal contracts has remained a critical challenge. This research explores how modern machine learning models like Legal T5, and Law GPT can automate contract analysis and improve review accuracy. Legal T5 summarizes the complex legal documents while Law GPT captures the critical risks. K-Means clustering was used to improve the contract segmentation by grouping similar clauses for a structured risk assessment. Tested on the CUAD dataset, the system's text summarizing findings showed coherence and a strong semantic similarity of 0.78, ensuring that the final summaries retained significant legal importance. The model achieved an average precision of 0.85, recall of 0.77, and F1-score of 0.80 across various risk categories. The findings prove that Legal T5, along with clustering, improves contract analysis with the ability to provide a context and structure-aware summary with better identification of risks.