German court rulings dataset helps AI identify legal interpretation styles
Classifying Interpretive Canons at the Sentence Level: A Benchmark from the German Federal Constitutional Court
Computation and LanguageComputers and Society
Summary
It is hard for AI language models to understand how judges reason through legal cases. The authors created a dataset of German court decisions broken down sentence by sentence, labeling different styles of legal interpretation. They tested several AI models to see how well they can tell these styles apart. The AI did okay, doing best with simpler grammatical interpretations and worst with more complicated systematic ones. Their expert-written prompts worked about as well as computer-optimized prompts, showing experienced guidance matters.
What this means in practice
- •For legal tech developers: Improve AI tools that analyze judicial decisions by enabling sentence-level classification of legal interpretation methods.
- •For legal document analysts: Support detailed review of court rulings by identifying interpretive canons in sentences automatically.
Authors
Felix Ringe
Abstract
Judicial reasoning remains challenging for large language models (LLMs) to analyze. This paper contributes a sentence-level benchmark for evaluating the ability of LLMs to classify interpretive canons as articulated by Larenz in the tradition of Savigny. Our contributions are threefold. First, we operationalize this conception of interpretation as classification criteria. Second, we provide a dataset of decisions of the German Federal Constitutional Court annotated at the sentence level. Third, we report baseline evaluations of four LLMs from three model families under expert hand-written prompts, compared against prompts optimized with Genetic-Pareto (GEPA). Mean F1 over the seven binary subtasks clusters between 70.4 and 79.2 across models, with grammatical interpretation usually the easiest canon to identify and systematic interpretation usually the hardest; under the tested configuration, GEPA-optimized prompts do not systematically outperform the hand-written ones, suggesting that the expert prompts provide a meaningful baseline.