Bye-bye, Bluebook? Automating Legal Procedure with Large Language Models
Matthew Dahl · 2025
Read Bye-bye, Bluebook? Automating Legal Procedure with Large Language Models by Matthew Dahl (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2505-02763.txt.
Description
Matthew Dahl and Eric Martínez test how well AI models handle one of law's most hated chores: formatting citations by the Bluebook rulebook. Their 2025 article gives the first measured check of model skill at this task. Citation rules are strict and fussy, so they make a clean test of rule following.
The team built a benchmark of 2,058 citation questions. Top models produced a fully correct citation only about 43 percent of the time. Even a reasoning model scored below human candidates in law journal editor tryouts. Handing models the rules helped little. A hybrid system, where the model parses and a fixed engine formats, did far better, reaching about 86 percent.
Catalog record
| CREATOR | Matthew Dahl |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 66 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2505-02763