About LegalArxiv
A global home for legal AI research. Find papers, understand their evidence and limitations, and follow the connections between them.
How we select research
We look for a clear contribution to legal AI, identifiable authors and sources, a method that can be examined, and limitations that are stated honestly. Coverage spans jurisdictions, languages and research traditions; the collection is still selective.
University and company research teams are welcome, as are independent researchers. Affiliation provides context. It does not replace evidence or determine admission on its own.
Each paper entry links to its source and explains what the work contributes, where its conclusions stop, and how it relates to other research.
What the entries and checks mean
Paper summaries are prepared with agent assistance and have not received independent scholarly review. Listing a paper does not endorse its findings.
Published research checks state their exact scope, method, artifacts and limits. A dataset check is not a reproduction of model performance. Indexing, editorial assessment and reproduction remain separate activities.
Original papers remain at their source. Versioned archive records carry their own provenance, rights, evidence and assessment status. Machine interfaces expose the public index and records with those distinctions intact.
Who maintains it
Maintained with heart by Jonas Hertner. LegalArxiv is an independent research project and is not affiliated with arXiv or the institutions whose research it indexes.
To contribute research or a correction, use the submission form. Contributions enter a private review process before any publication.