A new AI model developed by researchers at Stanford University demonstrates significant improvements in processing and summarizing lengthy legal documents. The system, named LegiSum, uses a novel architecture that better understands complex legal terminology and the hierarchical structure of contracts and case law. Initial tests show it can reduce a 100-page document to a concise, accurate summary in under a minute, with a reported 95% accuracy rate in identifying key clauses and obligations. The developers highlight its potential to increase efficiency for law firms and in-house legal teams, though they note it is designed as an assistive tool and not a replacement for attorney review. The full research paper detailing the model’s architecture and training data will be published next month. Read the full article at https://technologyreview.com/2024/05/15/legisum-ai-legal-documents/