MAUD: An Expert-Annotated Legal NLP Dataset for Merger Agreement Understanding

Steven H. Wang, Antoine Scardigli, Leonard Tang, Wei Chen et al.

Authors & affiliations

Steven H. Wang, Antoine Scardigli, Leonard Tang, Wei Chen, Dimitry Levkin, Anya Chen, Spencer Ball, Thomas Woodside, Oliver Zhang, Dan Hendrycks

Affiliations reported at publication:

ETH Zürich
Harvard University
The Atticus Project
The Nueva School
University of Wisconsin–Madison
Yale University
Stanford University
UC Berkeley

2023EMNLP 2023, pages 16369–16382

Read the paper Publication source

What it adds

Tests expert-annotated interpretation of merger agreement provisions through standardised deal-point questions.

Read with this limitation

The primary task supplies extracted clauses and covers selected large US public-company acquisitions closed in 2021. It does not measure the complete contract-review workflow or establish transfer to other transaction types.

Jurisdiction tags: United States
Language: English

Research artifacts

Resources linked by the authors.

Official code and data

Follow the source

Selected statements located in the linked primary sources. These results have not been independently reproduced here.

The primary task asks 92 standardised questions about extracted deal-point text from merger agreements.
1 Introduction
The paper's 152 agreements concern selected US public-company acquisitions exceeding $200 million that closed in 2021; smaller and private-company transactions are excluded.
6.3 Limitations

CUAD: An Expert-Annotated NLP Dataset for Legal Contract Review

Both provide expert-annotated contract resources; MAUD focuses on interpreting deal points rather than primarily locating clauses.

ContractNLI: A Dataset for Document-level Natural Language Inference for Contracts

Both test interpretation of contractual text through structured questions, using different agreement types and answer formats.

Entry status, source edition & rights

Index entry: Source-linked summary, prepared with agent assistance. Independent scholarly review is pending.
Archive admission: This bibliographic entry does not confer admission to the versioned archive.
Edition inspected: 10.18653/v1/2023.emnlp-main.1019
Reading coverage: Published EMNLP PDF: task construction, experiments, section 6.3 limitations, and Appendices A.12–A.14 inspected. Names follow the PDF, which prints Steven H. Wang and Dimitry Levkin; Anthology metadata differs. Linked code/data inspected, not executed.
Rights: External references only. Papers and research artifacts remain with their original hosts; their own terms apply.
Entry updated: 2026-10-15

Source-linked summary · independent assessment pending.