Critical AI Challenges in Legal Practice: An application to French Administrative Decisions

Khaoula Naili

Working Papers from HAL

Abstract: We use AI methods to evaluate the accuracy of several standard machine learning models for predicting judicial decision outcomes. We highlight the key steps and challenges in predicting judicial outcomes by applying these models to a database of administrative court decisions.These findings significantly contribute to our understanding of the potential advantages of AI in the context of predictive justice. We utilize AI methods to analyze administrative court decisions sourced from the database provided by the French Council of State. This analysis has been made possible due to the Council of State's decision to make its decisions publicly accessible since March 2022. Our innovative approach pioneers the use of prediction models on the open data from the French Council of State, addressing the complexities associated with data analysis. Our primary objective is to assess the accuracy of these models in predicting outcomes in French administrative tribunals and identify the most effective model for forecasting administrative tribunal court decisions. The selected models are trained and evaluated on multi-class datasets, where decisions are traditionally categorized into various classes.

Keywords: Artificial intelligence AI; Machine learning; Natural language processing; Predictive justice (search for similar items in EconPapers)
Date: 2023-11-30
Note: View the original document on HAL open archive server: https://univ-fcomte.hal.science/hal-04316581v1
References: Add references at CitEc
Citations:

Downloads: (external link)
https://univ-fcomte.hal.science/hal-04316581v1/document (application/pdf)

Related works:
Working Paper: Critical AI Challenges in Legal Practice: An application to French Administrative Decisions (2023)
Working Paper: Critical AI Challenges in Legal Practice: An application to French Administrative Decisions (2023)
This item may be available elsewhere in EconPapers: Search for items with the same title.

Export reference: BibTeX RIS (EndNote, ProCite, RefMan) HTML/Text

Persistent link: https://EconPapers.repec.org/RePEc:hal:wpaper:hal-04316581

Access Statistics for this paper

More papers in Working Papers from HAL
Bibliographic data for series maintained by CCSD ().