Our Research
We create open-source datasets and benchmarks to advance research in legal AI. Working with expert lawyers, we design and label real-world legal tasks and publish papers that help the community build smarter, more reliable legal AI systems.
CUAD is an expert-annotated dataset that helps AI spot the most important parts of legal contracts. Built from 500+ contracts with 13,000+ labels across 41 clause types, it teaches models to highlight key terms, speeding up contract review and making it easier to catch red flags.
MAUD is an expert-annotated dataset that helps AI understand merger agreements. Built from the ABA’s Public Target Deal Points Study, it turns real “deal points” into clear questions and answers, giving legal-tech tools a reliable way to read and interpret key terms at scale.
ACORD is the first expert-annotated dataset built to help AI find the right legal contract clauses. It includes real lawyer-written queries and thousands of rated clauses, making it easier to draft and review complex contracts like Limitation of Liability and Indemnification.
